United States       Office of Research and    EPA/600/R-96/090 i^
          Environmental Protection    Development       September 1996
          Agency          Washington DC 20460



&EPA    Problems of Aquatic


          Toxicology, Biotesting and


          Water Quality Management

-------
                                           EPA/600/R-96/090
                                             September 1996
<0
    Problems of Aquatic Toxicology,
           Biotesting and Water
           Quality Management

Proceedings of USA-Russia Symposium,
           Borok, Jaroslavl Oblast,
               July 21-23,  1992

                 Editorial Coordination By
                Richard A. Schoettger, Ph.D.
                 Midwest Science Center
                 National Biological Survey
               U.S. Department of the Interior
                  Columbia, MO 65201
                     Published by
               Ecosystems Research Division
                   Athens, GA 30605                 ft  y
                              M s Environmental now
            National Exposure Research Laboratory
             Office of Research and Development
             U.S. Environmental Protection Agency
              Research Triangle Park, NC 27711
                                          Printed on Recycled Paper

-------
                                  Notice
      The views expressed in these Proceedings are those of the individual authors
and do not necessarily reflect the views and policies of the U.S. Environmental
Protection Agency (EPA).  Scientists in EPA's Office of Research and Development
have prepared the EPA sections, and those sections  have been reviewed in
accordance with EPA's peer and administrative review polices and approved for
presentation and publication.

-------
                                Foreword


      For more than two decades, cooperation and exchange of scientific information
under the USA-(former) 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 addressed under the joint Agreement's Working Group on
Cooperation in the Area of Water Pollution Prevention.  These are Project 02.02-11
"Planning and Management of Water Quality in River Basins," 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 the papers that were
presented at the symposium held in Borok, Jaroslavl Oblast, in July 1992.

                                          Rosemarie C. Russo, Ph.D.
                                          Director
                                          Ecosystems Research Division
                                          Athens, Georgia

-------
                                 Abstract
      Sixteen papers delivered at the international symposium entitled "Problems of
Aquatic Toxicology, Biotesting and Water Quality Management" are presented. The
impact of permethrin on zooplankton in high latitude ponds is examined, and an aquatic
food chain model for simulating toxicant impact on zooplankton population dynamics in
a water body is presented. Problems related to monitoring and modeling pollution from
agricultural watersheds are described, and a water quality planning model for
agricultural watersheds (Answe'rs-2000) is presented. The effects of metals on fish are
investigated in studies of mercury content and ultrastructure of gills and scales offish
in Russian lakes that are polluted by atmospheric deposition and of beryllium damage
to gills in three freshwater fish species. In another study, the effect of heavy metals
and chlorpyrifos on a continuous-flow mesocosm is examined.  An investigation of
relationships between waterborne tetrachloroguaiacol uptake and oxygen consumption
by fishes is described. Management issues are addressed in reports on a Green
Bay/Fox River (Wisconsin) mass balance study, on an effort to restore aquatic
resources to the lower Missouri River,  and on a project to improve habitat in riverine
lakes through hydrologic modification.  A classification scheme for wetlands in Russia
is introduced.

       The use of simulation models to investigate water quality in the lower Don
River is described. Complex anthropogenic factors that  influence the ecological state
of the upper Volga River are addressed; physiological and biochemical indicators of
contaminant stress in aquatic organisms of large river systems are identified.  Attention
is directed to the work of N.V. Lazarev in the early development of structure-activity
relationships for environmental toxicology, leading to modern approaches to assessing
contaminant effects on man and on organisms in the environment.
                                       IV

-------
                                Contents
Foreword  	iii
Abstract  	iv

Simulating Toxicant Impact on Zooplankton Population Dynamics in a Water Body . . 1
      Y.M. Plis andR.R. Lassiter

Green Bay/Fox River Mass Balance Study-Management Implications	 9
      LF. Wible, J.G. Konrad and J. Steur

Problems Related to Monitoring and Modeling Pollution from Agricultural
      Watersheds	 33
      V.Z. Kolpak, T.A. Dillaha III, D.B. Beasley, G.Y. Platonova and AS. Voronkin

Answers-2000: A Water Quality Planning Model for Agricultural Watersheds	 43
      T.A. Dillaha III,  Vladimir 2.  Kolpak, F. Bouraouri, D.B. Beasley and
      G. Y. Platonova

Relationships Between Waterborne Tetrachloroguaiacol Uptake and Oxygen
      Consumption by Fishes	 54
      J.F. Neuman, R.V. Thurston, C.J. Brauner and D.J. Randall

An Investigation of the Lower Don River Water Quality Using Simulation Models ... 64
      M.G. Yereschukova, E.Z. Hosseinipour, R.B. Ambrose, V.V. Tsirkunov,
      R.C. RussoandA.M. Nikanorov

Early Development of Structure-Activity Relationships for Environmental
      Toxicology in Russia	 77
      R.L Lipnick, V.A. Filov and R.C. Russo

Mercury Content and Ultrastructure of Gills and Scales of Fish from Lakes
      in North and Northwestern Russia That Are  Polluted by Atmospheric
      Deposition 	 90
      T. Haines, V. Komov,  C. Jagoe and V. Matey

Mortality and Gill Damage from Beryllium in Acidic Water: A Comparison of Acute
      and Chronic Responses in Three Freshwater Fish Species	  109
      V.E. Matey, C.H. Jagoe, T.A. Haines, V.T. Komov and L. Cleveland

Impact of Permethrin on Zooplankton in High Latitude Experimental Ponds	  129
      T.A. La Point, O.D. Zhavoronkova and D.F. Pavlov

-------
The Effect of Heavy Metals and Chlorpyrifos, Separately and in Combination,
      on a Continuous Flow Mesocosm Aquatic System 	 148
      G.A. Vinogradov, F. Stay, P.P. Umorin, A.S. Mavrin, A.K. Klerman
      E.I. Koreneva, S.A. Kurbatova, 1.0. Solntseva and G.I. Vinogradova

Biochemical and Physiological Indicators of Contaminant Stress in Aquatic
      Organisms of Large River Systems	 162
      D.R. Buckler and D.E. Tillitt

Classification of Wetlands in Russia 	 175
      A.M. Nikanorov, A.V. Zhulidovand W. Sanville

Ecological State of the Upper Volga River As Influenced By Complex
      Anthropogenic Factors	 183
      N.P. Smirnov and A.I. Kopylov

Restoring Aquatic Resources to the Lower Missouri River: Issues
      and Initiatives	 192
      D.L.  Galat, J.W. Robinson and L.W. Hesse

Hydrologic Modification To  Improve Habitat in Riverine Lakes:  Management
      Objectives,  Experimental Approach and Initial Conditions  	 239
      B.L. Johnson, J.W. Barko, Y. Gerasimov, W.F. James, A. Litvinov,
      T.J. Naimo, J.G. Wiener, R.F. Gaugush, J.T. Rogala and S.J. Rogers
                                      VI

-------
                 SIMULATING TOXICANT IMPACT ON ZOOPLANKTON
                     POPULATION DYNAMICS IN A WATERBODY

                                        by

                             Y. M. Plis1 and R. R. Lassiter2


                                    ABSTRACT

       In the past 20 years models have become useful tools for studying processes in
waterbodies. We developed an aquatic food chain model.  It was used to represent
zooplankton population dynamics in a waterbody, through seasonal changes in water
temperature and growth dynamics.


                                  INTRODUCTION

       Over the last 20 years, ecosystem models have become effective tools for
investigating the complex physical, chemical, and biological processes in waterbodies. Most
such models have represented aquatic organisms in all ecological roles simply as scalar
biomass concentrations.  Their approach assumes that differences among individuals in a
population can be ignored.  This approach is acceptable if aquatic ecosystem models are
used for investigation of the processes of eutrophication in waterbodies, when the primary
interest is in exploration of the dependence of plankton biomass changes on enrichment of a
waterbody by nutrients. However, greater resolution of the intrapopulation structure is
needed if one is to understand population dynamics in other situations.  For example,
competition for resources; predator avoidance; and, especially, ability to withstand severe
environmental stress by toxicant impact depend on an individual's physiological state.

       Models of physiologically structured populations were discussed in Metz and
Diekmann  (1986). Physiologically structured models that include body composition elements
as variables of their  individual model component have  been proposed and used by Hallam et
al. (1988a, b, 1990a, b).  Such equations are often referred to as "McKendrick-von Foerster"
equations, a general form of which is
Ukrainian Scientific Center for Protection of Waters, Kharkov, Ukraine.
2U.S. Environmental Protection Agency, Environmental Research Laboratory, Athens, GA
 (USA).

-------
                                  (f,  P)P(f,
where P is the population density function, t is time, a  is a vector of individual physiological
properties (age, masses of organism body components, or contained substances), g; is the
growth rate for the ith physiological property, and M is the mortality rate function.

       In this work, we combined equation 1 with equations from the  individual Daphnia
model of Hallam et al. (1990a), equations from the model of toxicant assimilation by
zooplankton of Hallam et al. (1988a), and transport equations derived by Plis (1981) using a
curvilinear coordinate system. The combined model has been used to represent zooplankton
population dynamics in a waterbody whose environment varies in time and space.  These
variations include seasonal changes in water temperature and growth dynamic of the
phytoplankton that is used as food by zooplankton-all in a spatially variable field of water
currents.

       Hence, toxicant effects on temporal-spatial variations of the zooplankton population
will be estimated as a result of the solution of four related problems: 1) transport of dissolved
and solid components by water currents and turbulent diffusion; 2) ecosystem interaction of
different trophic levels of the model-zooplankton population, phytoplankton, and nutrient; 3)
assimilation of toxicant by phytoplankton and zooplankton organisms; and 4) population
zooplankton dynamic and growth of zooplankton individuals under the conditions of lethal and
sublethal toxicant impacts.
                  COORDINATE SYSTEM AND TRANSPORT MODEL

      The transport model uses curvilinear coordinates representing the spatial boundaries
of waterbodies as coordinate surfaces. These coordinates offer the advantage of working
with a simple waterbody shape and of using a mesh with the optimal numbers of nodes for
approximating the waterbody area with an acceptable level of accuracy.  The procedure for
calculating curvilinear coordinates was described by Plis (1992).

      The transport equation for a shallow water body has the following form
                     dC+
                     dt
                                                                                  (2)
                                  O
where C is the matrix-column, elements of which are the dependent variables of the model; F
is the matrix-column of functions of the chemo-biological interactions of model components;
Q is the matrix-column of external sources of model components; J=d (x, y) / d(f, rj) is the
Jacobian of the transformation from Cartesian coordinates (x, y) to curvilinear (f, 77); vf and v"

-------
are the contravariant components of current velocity; v°g is the component of sediment
velocity; DL is the horizontal coefficient of turbulent diffusion; and Sf and S" are components
of the full stream vector.

      The system of equations of the Ekman theory of marine currents expressed in the
curvilinear coordinate system was used for calculation of current components, Sf and S".
                MODEL FOR DYNAMICS OF AN INDIVIDUAL DAPHNID

       A mathematical model describing life history and bioenergetics of an individual
daphnid, along with tables of parameters, was presented by Hallam et al. (1990a). Following
this model, an individual daphnid is represented as consisting of two body components-
structure and lipid.  Its dynamics are represented by the rates of change of these two
components,  as controlled by energy supply and demand.  Input of structure and lipid from
food balanced against the demands of maintenance, carapace formation, work, and
reproduction determines the net rate of change of the body components.  Reproduction and
carapace formation,  in contrast to the other demands, are computed on a discrete time
interval.  In general,  we follow this model closely, including the use of most of its parameters.

       To use the model in a dynamic situation including seasonal dynamics and a wide
range of food densities, however, we had to make some alterations.  Specifically, we added
the representation of the  influence of water temperature on growth and mortality and a
limitation on maximum size of the protein and lipid compartments.
                     MODEL FOR PHYTOPLANKTON DYNAMICS

       Phytoplankton is primary food for daphnids, and in this model phytoplankton dynamics
depend on availability of mineral nutrients, on changes in temperature, and on consumption
by zooplankton. Our nutrient-phytoplankton model is structured only with respect to space
and is the simplest that we could conceive to provide dynamics to support a physiologically
and spatially structured population of zooplankton.
               MODEL OF TOXICANT ASSIMILATION BY ZOOPLANKTON
                              AND PHYTOPLANKTON

      We assumed that the dissolved toxicant is assimilated by individuals of the population
through two mechanisms. The first is penetration across the surface of the organism's body
by diffusion.  The second is toxicant accumulation from  food. In accordance with our
assumptions, the food for zooplankton is phytoplankton.  It is assumed that processes of
toxicant accumulation by phytoplankton can be described using an equilibrium hypothesis.
Removal of toxicants from the zooplankton organism takes place with nonassimilated food
and through diffusive exchange across the body surface of the organism.

      We assume here that the toxicant does not impact on phytoplankton life processes.
In relation to zooplankton organisms, however, the toxicant slows down processes of
                                      -  3

-------
accumulation and consumption of energy in accordance with theoretical suppositions
described below.  If the toxicant concentration in the liquid phase of the organism exceeds
values equal to the lethal concentration (LC50), it is assumed this organism will die.

       It was assumed also that the distribution of toxicant in the waterbody depends only on
conditions of transport by currents and turbulent diffusion, and the location and intensity of
external sources of pollution.  Biochemical and physical processes of toxicant transformation
in a waterbody have not been taken into account in the present work.

       The equation of the toxicant balance in the individual organism of Daphnia M. (Hallam
et al. 1988a) was used as the basis of the model of toxicant assimilation. This equation was
modified to take into account losses of toxicant caused  by reproductive processes. The
equation can be written as
                        dmT
                       -    = SKW (Cw -Ca YFCF -ECE -ff,
where mT is the mass of toxicant accumulated into an individual organism, mg; Kw is the
coefficient of conductivity across organism body surface, mm/day; S is the effective surface
area of the organism body, permeable for chemical, mm2; Cw and Ca are the concentrations
of toxicant in the ambient water and in the aqueous phase of the organism, mg/L; F is the
mass of food eaten per unit time, mg/day; E is mass of material defecated per unit time; CF
and CE are concentrations of toxicant in the food and in the feces; R is mass lost for
reproduction per unit time, mg/day.

      Using the assumption of internal equilibrium between concentrations of the toxicant
assimilated in the three basic phases of the organism (aqueous, lipid, and structural [nonlipid
organic]), we can represent  the concentration of toxicant in the aqueous phase Ca, in
accordance with Hallam et al. (1988a) and Burns et al. (1990) by the expression

                                            rri-r                                    IA\
                              r.  = 	L	                             (4)
                                    ~* is. s.-  is is s. \
                                    mL(t>2c>  +/V+AS o2)
where mT = (maCa + mLCL + msCs); ma, n\, and ms are mass of aqueous phase, mass of
lipid, and mass of structure, respectively, in the body of the individual; 62 = ms/n\; 6 = ms/ma;
KL and Ks are thermodynamic partition coefficients of the toxicant between the lipid and the
water, and the structure and the water, respectively; CL and Cs are concentrations of toxicant
assimilated in lipid and structure phases of the organism, respectively.

       In the framework of the present model, the assimilation of a toxicant by phytoplankton
was taken into account using the assumption about equilibrium between ambient toxicant
concentration and toxicant concentration in phytoplankton biomass.

                           CL  = KLCP fL,   C5  = Ks Cp /s ,

-------
where Cp is the phytoplankton biomass, mg Chlorophyll a/L; fL and fs are fractions of the lipid
and the structure of the phytoplankton biomass.

       For term R (equation 3) we assume that the loss of individual mass during the
reproductive process is accompanied proportionally by the loss of the toxicant mass
accumulated in the parent organism's body, or

                                    e        -1    e                               (6)
                                 rrij - mT mL  mL,                               v '


where, m? and  m^ are mass of toxicant and mass of lipid, respectively, in one newborn;
And R becomes

                                     R = Pn mf,                                  (7)

where Pn is the number of newborn in the one birth event of one organism of zooplankton.


                 MODEL OF TOXICANT EFFECTS ON AN INDIVIDUAL

      The problem of modeling lethal toxicant effects on Daphnia M. population dynamics
was discussed by Hallam et al. (1990b).  In their model, the amount of chemical in the
aqueous phase of the organism was compared with the LC50 determined on the basis of
qualitative structure-activity relationships (QSARs) that connect chemical property, e.g., the
octanol-water partition coefficient ( /G ) , with mortality of the individual.  We assumed that if
                                                   'w
the individual will die.

       The principal difference between lethal and sublethal effects, which was accepted as
the initial assumption in this paper, is that the sublethal reaction of the individual is a
monotonically increasing function of the internal concentration of the toxicant, Ca. This
function has a lower bound, where the sublethal effect is not observed, and an upper bound,
where the lethality occurs.  Lassiter (1990) proposed a model for the effect on organisms
exposed to a sublethal toxicant concentration, Cw.  With prolonged exposure, the organism's
internal aqueous concentration, Ca, approaches Cw at a  rate controlled by the organism's
exchange mechanisms (gill, gut), and by its body composition and circulation.
                     NUMERICAL REALIZATION OF THE MODEL

       The scheme of component-by-component splitting (Marchuk 1982) was used for the
numerical solution of the system of quasilinear equation 2. This method reduces a
complicated problem to a simpler one when the  initial differential operator of the system is
positive-semidefinite and can be represented as a sum of positive-semidefinite simpler
operators.

-------
       Besides a formal numerical scheme for solution of the system of equations, other
simplifications were made in the interest of computational feasibility. Although these
simplifications undoubtedly  reduced the fine-scale resolution and variance, they were made
subject to conservation constraints so that mass and numerical density (number of individuals
per unit volume of water) of the population were preserved.

       Initially, the population density function is a function of 6 variables

                             P= P (a, t, mv m2, my /7?4),                          ^

where a is the age of the individual organism; m^ m2, m3, and m4 are the masses of
individual  lipid, structure,  nonlabile structure,  and assimilated toxicant,  respectively. The
following assumptions were made to simplify  equation 1 .  The  daphnid population was shared
by I physiological clusters distinguished,  as in Hallam et al. (1990a), by different ratios of
maximal filtering rate to organism  length squared.  Each  of the I physiological clusters was
shared by N age groups. All individuals  of any age group of a physiological cluster are
physiologically and morphologically identical.  This is accomplished first by an assumption
about average properties of the first age group.  The first age group of a  cluster consists of
offspring from all reproductive age groups of  that physiological cluster.  Parents are
distinguished, in addition to age, by lipid, by structural (labile and nonlabile) body
composition,  and by assimilated toxicant. Hence, by the  model of Hallam et al.  (1988b,
1990b), they  produce offspring that also vary somewhat.  Our  assumption is that these
differences in the offspring (i.e., the first  age  group) are inconsequential, and  hence, we
average these properties to derive their values for the entire age group of the physiological
cluster. Additionally, populations are  mixed by spatial movement. Here we assume that the
properties of  each age group of a physiological cluster at a spatial point in a waterbody are
the weight averages of the remaining residents and the recruits from neighboring spatial
points.

       As a result of the foregoing assumptions, population  density and individual mass
characteristics for each ith physiological cluster can be simplified to

        p ' = P1 (a, t );  mj = m] (a,  t),     / = 1 ..... 4                            (10)

       This, together with assumed functions describing zooplankton population mortality,
permits equation 1 to be transformed to

                               dP     dP
                              ---—


            P'(0,t) =   fi(a,  t, ml m^ m£, m^P'(a, t)da,
where M'a and  M'p are the age dependent and density dependent mortality, respectively; £ is
the birth function that represents the number of neonates born to an individual of age a with
mass characteristics m^ m2, m3,  and m4 at time t (Hallam et al.  1990a).

-------
                              NUMERICAL EXPERIMENTS

       To explore the behavior of this model, we conducted numerical simulation of seasonal
dynamics of the nutrient-toxicant phytoplankton-daphnid system in a hypothetical, shallow
waterbody with arbitrary shapes for the lateral boundaries and with current velocities low
enough that they did not dominate ecological processes in determining concentrations of
nutrients or densities of phytoplankton and zooplankton.  We assumed a time-invariant
distribution of current velocities during a simulated year.  Zooplankton population was
assumed to consist of only one physiological cluster (1 = 1) consisting of 50 age groups (N =
50) with the age group interval being 1  day.

       Although our computations  relate to a hypothetical situation, we selected values of
physical properties of the hypothetical system to be well within the range of real physical
systems.  We discuss the  results of our simulations of the hypothetical situation for two
purposes: first,  to show the behavior of the Daphnia population model in the context of a
spatially and  temporally variable system; and second, to give a concrete description of the
nature of some of the results that can be expected from a spatially and physiologically
distributed aquatic food chain model.

       As we simulated the seasonal dynamics of the zooplankton population in one
hypothetical waterbody, various spatio-temporal distributions of population characteristics
were obtained.  These characteristics included  numerical densities, population biomass,
ratios of the population lipid to population biomass, and distributions of densities and mass
characteristics among age groups.  We noticed strong correlations between these
characteristics and the spatio-temporal  variations of phytoplankton biomass and concentration
of toxicant in the waterbody.  We found that the distribution of individual lipid content is the
most sensitive indicator of environmental stress.  For an individual organism, lipid fluctuates
more in response to environmental stress fluctuations than does body structure or total
biomass.  The next  most important characteristic  for a daphnid population  is the distribution
of age at first reproduction, which depends on the distribution of lipid and nonlipid mass
content.  It was shown that the inhibiting influence of toxicants on activity and reproduction
decreases with increasing  individual lipid content.
                                    CONCLUSIONS

       A spatially- and physiologically-structured aquatic food chain model was developed.  It
represents the transport of both dissolved and passively suspended components by water
currents and turbulent diffusion.  The model also represents the chemical-biological reactions
along a food chain, including chemical, nutrient, phytoplankton, toxicant, and age- and
mass-structured zooplankton populations. The structured population is represented by an
extended McKendrick-von Foerster equation that requires an individual model at the
physiological level of resolution for each of the dimensions of the population model (age,
lipid, structure, nonlabile structure, and toxicant body burden).

       A curvilinear coordinate system also  was employed to improve computational
efficiency in  comparison with an ordinary Cartesian coordinate system. This approach
improves computational efficiency by permitting the waterbody to be represented in the

-------
transformed coordinates as a simple rectangular shape, and to be approximated by a grid
with the optimal number of nodes for numerical approximation at an acceptable level of
accuracy.  The algorithm for numerical solution of the model equations was developed using
the method of component-by-component splitting.

       The example of the model application to a polluted lake showed opportunities to
obtain additional comparative information for decision makers about the behavior and life
conditions of populations in clean and polluted areas of waterbodies. The success of this
and other attempts depends on existing biologically and ecologically realistic individual based
models.
                                   REFERENCES

Burns, L. A., M. C. Barber, S. L. Bird, F. L. Mayer Jr. and L. A. Suarez.  1990. PIRAHNA:
       pesticide and industrial chemical risk analysis and hazard assessment.  Version 1.0.
       U.S. Environmental Protection Agency, Athens, GA.  (Unpublished report.)

Hallam, T. G., R. R. Lassiter and S. A. L. M.  Kooijman.  1988a. Effects of toxicants on
       aquatic populations. In S. A. Levin, T. G. Hallam and L. J. Gross (eds.), Mathematical
       ecology. 11. Applications.  Springer-Verlag, New York.

Hallam, T. G., R. R. Lassiter, J. Li and W. McKinney.  1988b.  Physiologically structured
       population  models in risk assessment. In Luigi Ricciardi (ed.), Biomathematics and
       related computational problems.  Reidel, Hingham, MA.

Hallam, T. G., R. R. Lassiter, J. Li and L. A. Suarez.  1990a. Modelling individuals employing
       an integrated energy response: application to Daphnia. Ecology 71:938-954.

Hallam, T. G., R. R. Lassiter, J. Li and W. McKinney.  1990b.  Toxicant-induced mortality in
       models of Daphnia populations. Environ. Toxicol. Chem. 9:597-621.

Lassiter R. R.  1990. A theoretical basis for  predicting sublethal effects of toxic chemicals.
       U.S. Environmental Protection Agency, Athens, GA.  (Unpublished report.)

Marchuk, G. I.  1982.  Methods of numerical  mathematics.  Springer-Verlag,  New York.

Metz, J. A. J. and O. Diekmann. 1986.  The dynamics of physiologically structured
       populations.  In Lecture notes in biomathematics, vol. 68. Springer-Verlag, Berlin.

Plis, Y. M.  1981.  Model of transport and transformation of dissolved substances in water
       body for choice of water protection measures. Pages 96-108 in Complex water
       protection measures.  Kharkov, Ukraine.

Plis, Y. M.  1992.  An approach to calculating wind-driven currents  and transport of
       substances in unstratified water bodies using curvilinear coordinates.  Water Resour.
       Res.  28:83-88.
                                         8

-------
                GREEN BAY/FOX RIVER MASS BALANCE STUDY-
                          MANAGEMENT IMPLICATIONS
                                          by

                   Lyman F. Wible1, John G. Konrad2, and Jeffrey Steur3



                                     ABSTRACT

       In 1986, the Green Bay/Fox River Mass Balance Project was initiated to demonstrate the
feasibility of the mass balance concept as a water quality management tool.  The project
monitored and refined mathematical models which represented the inputs, transformations, and
outputs of four toxicants in the waters of Green Bay, Wisconsin. Green Bay is an embayment of
Lake Michigan and historically has received significant pollutant loads from the industries and
municipalities along the Lower Fox River. The toxicants of concern for the Green Bay/Fox River
mass balance study are polychlorinated biphenyls (PCBs) at the congener level, dieldrin, lead and
cadmium. Existing hydrodynamic models, water/sediment models, and a food chain model were
coupled to estimate the pollutant body burden in target fish species, mainly carp, brown trout, and
walleye.  Preliminary results of the modeling and the implications of the model results are
presented. Project management experiences and lessons learned from conducting a large multi-
participant project also are discussed.
'Division of Environmental Quality, Wisconsin Department of Natural Resources, Madison, WI
53707, USA
2Bureau of Research, Wisconsin Department of Natural Resources, Madison, WI 53707 USA
'Bureau of Water Resources Management, Wisconsin Department of Natural Resources,
Madison, WI 53707, and U.S. Geological Survey, Madison, WI 53717, USA

-------
                                 INTRODUCTION
    This paper presents some of the preliminary results from the Green Bay/Fox River
Mass Balance Project, which is an  analysis of the sources, sinks, pathways, and fates of
four pollutants in the Green Bay and Fox River aquatic systems.  Green Bay is a part
of Lake Michigan, one of the Great Lakes, and is the largest freshwater estuary in the
world. The Bay is 4,520 km2 (1,640 mi2) in areal extent, with an average  depth of 3-5
m (10-15 feet) in the southwestern end and 15 m (49 feet) in the northeastern end with
a maximum depth of 54 m (177 feet).  Figures  1 and 2 depict the location of Green
Bay and the Fox River. The Fox River is the major tributary to the Bay,  and extends
63 km (39 mi) from Lake Winnebago to the river mouth at Green Bay. The City of
Green Bay and the other communities along the Lower Fox River are home to 270,000
residents, or 6% of the population of the State.  This area includes 20 pulp and paper
mills discharging either directly to  the Fox River or indirectly via municipal sewage
treatment plants, and is generally considered the largest concentration of pulp and
paper mills in the world.

    Because the project is just entering its final stages, all results are not yet available.
Some of the preliminary results are presented in this paper. Additional results will be
reported by project investigators in future papers.  The status and general findings for
the entire study  area are included  in this paper, but the focus is primarily  on the lower
Fox River, for which the earliest project  results have become available.

    The Green Bay/Fox River  Mass Balance Study began in 1986 after more than one
yr of dialogue among the various government and private organizations with  an interest
in these waters.  The study has involved  more than 20 major organizations and
laboratories of the U.S. Environmental Protection Agency. The participating
organizations are listed in Table 1. The  direct expenditures in this project have totalled
more than $10,000,000 with substantial indirect  expenditures by the participants and
other supporting organizations  through their efforts to support, assist, and make this
historic study effective.

    During the first 2 yr,  1986 and 1987, project participants developed monitoring and
quality assurance programs and began field testing analytical and monitoring methods.
In 1988 the field work was initially tested and sampling started with a shakedown cruise
on October 1988.  Intensive field work was initiated in the spring of 1989  and continued
through May 1990.  From 1990 to 1992,  samples were analyzed, data interpreted, and
modeling was conducted.  By October 1992 the modeling is scheduled for  completion,
while the management analyses using these sophisticated models may be expected to
extend for many additional years.  A final report on the  project is due at  the end of
calendar year 1992.

                                        10

-------
          woor
                       era
                                    srx
45%
45V
45-W
tee
44'W
                                                  87*6'



                                                   Kkmlen
                                                               87W
                                                                            86'45'
                                                                                         86'W
               GREEN  BAY
               Band on NOAA Chart 14902
Water Column
Sampling Stations
                                                                                                      46W
                                                                                                      45'45'
                                                                                                      45'W
                                                                                                      45-B'
                                                                                                      45W
                                                                                                      W4S
      •SFOX R.
        88W
                                    tno
                                                 tire
                                                                              Green Bay Vass balance Study
                                                                             86-45'
                  Figure 1.  Location of Green Bay and water column sampling stations.
                                                  11

-------
                                                                            Green Bay
                                                                      De Pere
Little Lake
Butte Des Mortes
•
Paper Mill
               Figure 2.  Location of the Lower Fox River and point sources sampled.
                                           12

-------
     TABLE 1.  MAJOR COMPONENTS
                           PROJECT AND
OF THE GREEN BAY/FOX RIVER MASS BALANCE
RESPONSIBLE ORGANIZATIONS
  Major Components
                    Responsible Organization
-Fox River transport modeling (above DePere)
-Point source monitoring
-Urban nonpoint source estimates
-Ground water impacts
-Abandoned landfill impacts
-Fish collection and processing
-Project management
-Lower Fox River deep core sediment samples

-Conventional parameter analyses (Tributaries and
Fox River above DePere)
-PCB analyses in Fox River sediment and Fox River
water samples above DePere

-Green Bay water column monitoring
-Project management
-Conventional parameter analyses (Green Bay and
Lower Fox River)

-Modeling coordination
-Fish and other biota analyses
-Lower Fox River transport modeling (below
DePere)

-Water column  analyses
-Analysis of plankton samples

-Food chain modeling
-Statistical integration

-Green Bay mass balance modeling
-Physical characterization of Fox River bottom
sediments
-Bottom sediment resuspension and flux

-Desorption kinetics
-Sedimentation and volatilization rates
-Water volume transport
-Sediment flux in Green Bay

-Atmospheric monitoring
-Atmospheric deposition
-Quality assurance/quality control
-Bioaccuumulation of PCB in phytoplankton

-Local area coordination

-Tributary monitoring
-Sediment mapping

-Data management
        Wisconsin Department of Natural Resources - Madison and Green
        Bay, WI
        Wisconsin State Laboratory of Hygiene - Madison, WI
        U.S. Environmental Protection Agency - Great Lakes National
        Program Office-Chicago, IL
        U.S. Environmental Protection Agency - Large Lakes Research
        Station-Grosse He, MI
        U.S. Environmental Protection Agency - Environmental Research
        Laboratory-Duluth, MN

        Manhattan College - New York, NY
        University of Notre Dame - South Bend, IN
        State University of New York-Buffalo, NY
        Clarkson University - Potsdam, NY
        Limno-Tech, Inc. - Ann Arbor, MI
        Youngstown State University - Youngstown, OH

        University of California-Santa Barbara, CA
        University of Wisconsin-Madison, WI
        University of Wisconsin-Milwaukee, WI
        Wisconsin Sea Grant Program - Madison, WI

        National Oceanic and Atmospheric Administration - Ann Arbor, MI
        Illinois State Water Survey - Champaign, IL
        DePaul University - Chicago, IL
        University of Minnesota-Minneapolis, MN

        University of Minnesota-Minneapolis, MN
        University of Wisconsin-Green Bay, WI

        U.S. Geological Survey - Madison, WI


        Computer Sciences Corporation - Grosse He, MI
                                                       13

-------
                ENVIRONMENTAL PROBLEMS IN GREEN BAY
    Eutrophication problems in the Great Lakes have focused remedial programs on
reduction of phosphorus and other conventional pollutants.  Current research on
environmental problems in the Great Lakes indicates persistent toxic substances are a
major problem (1).  Green Bay, as a major arm of Lake Michigan, constitutes an
important subsystem of the whole lake.  The predominant water quality problems in the
waters of Green Bay have been historically associated with the Green Bay metropolitan
area and the discharge of the Fox River.  Green Bay and the Fox River have a long
history of water quality problems derived from the contributions of pulp  and paper
mills, municipal sanitary wastes, and  other sources (2).

    From the 1930s to the 1970s dissolved oxygen conditions grew worse.  The
biochemical  oxygen demand (BOD) load discharged to the Fox River steadily increased
because of paper industry growth and - to a lesser extent - population growth.  The
results of pollution abatement efforts since 1972 have been  excellent and have shown
dramatic reductions in BOD and suspended solids. The discharge of conventional
pollutants (BOD) to the Fox River has been reduced by more than 95%, from 375,000
pounds/day in 1971 to 17,400 pounds/day in 1984 (2,3).  The biological response in the
aquatic system has been equally dramatic.  The waters of the  Lower Fox River and
Green Bay, largely unused from the  1930s to the 1970s for recreation or fishing, have
become a major attraction and economic asset. A highly productive walleye fishery has
returned, land uses are changing, marinas and  shoreland development are flourishing,
and commercial and sport fishing as  well as recreational boating have increased beyond
any expectations at the time of the cleanup. Community pride in the accomplishments
is evident.

    In spite of these achievements, monitoring programs using improved analytical
precision and information on the recovering fisheries and wildlife have indicated that
health advisories are prudent to guide the public concerning human consumption of fish
and wildlife.  Accordingly, the past decade has been a time of increased  public
sensitivity and awareness to governmental warnings on the potential health hazards of
food from these waters.

    In 1981, the International Joint Commission (IJC) identified 39 Areas  of Concern
(increased to 42 in 1985 and to 43 in 1991) in the Great Lakes Basin.  State and local
governments began to develop Remedial Action Plans (RAPs) in 1986 to address these
problem areas. Annex 2 of the 1987 Protocol to the 1978 Great Lakes Water Quality
Agreement formalized the review and approval of RAPs by the IJC.  The  persistent
environmental problems in Green Bay, coupled with the IJC's call for RAPs, prompted
the development of the RAP for the lower portions of Green Bay and the Fox River.

                                        14

-------
Supported by major public participation during its development, the Lower Green Bay
RAP was completed in 1988.  The Lower Green Bay RAP was the first prepared in the
Great Lakes Basin and first to receive the approval of the IJC. While public support
has been strong, public tolerance for scientific uncertainty has been fairly limited, with
local officials and citizens alike demanding as much scientific insight as possible.
Fortunately, an extensive database and understanding of the ecosystem of Green Bay
was  available from previous research studies.  For these reasons the Green Bay Mass
Balance Project was seen as a very positive and scientifically credible addition to the
RAP process.

                THE MASS BALANCE PROJECT STUDY DESIGN
    As discussed by Fink and Wise (1), the intent of the mass balance concept is to
evaluate the sources, transport, and fate of contaminants in the study area. The
approach requires that the quantities of contaminants entering the system, less
quantities stored, transformed, or degraded within the system, must equal the quantities
leaving the system. Toxicants of concern in the Green Bay/Fox River mass balance
study are Polychlorinated Biphenyls (PCBs) at the congener level, dieldrin, lead, and
cadmium. These toxicants were chosen for their ecological significance (PCBs) or as
representatives of  classes of contaminants (pesticides, methylating metals, and inorganic
metals).

    The initial mass balance study area was limited to  the waters of Green Bay with
the shoreline serving as the boundaries for the project.  The Wisconsin Department of
Natural Resources believed that because the lower Fox River played such an integral
role in the water quality dynamics of the shoreline waters and open waters of the Bay,
this portion of the river should  be included in the study area.  For this reason, a major
effort to model the transport of PCBs in the 51.5 km (32 mi) portion of the lower Fox
River from Lake Winnebago to the last dam (located at DePere, Wisconsin) was also
developed.  The overall project objectives are

          Quantify and map the presence of selected  pollutants within the aquatic
          system  and the  environmental compartments of the system.
          Identify the loads of these pollutants moving into and through the system.
          Identify the sources  of these pollutants.
          Identify the water/sediment/food chain interactions within these aquatic
          systems.
          Provide a tool  for the analysis of alternative management scenarios
          (addressing the sources of toxic substances and existing accumulations or
          inventories of the substances).

    Because of the extensive field work involved in  the project and the need to
coordinate the efforts of a large number of different organizations, project leaders
worked almost 2 yr to develop  the study design and evaluate appropriate field and

                                         15

-------
laboratory procedures. We targeted the field work to the characterization of the inputs,
the outputs, and the internal  interchange mechanisms for transfer of pollutants from
one environmental compartment to another inside the study boundary.  Figures 1 and 2
show the study area and  the location of the surface water column sampling stations and
point sources.  For this study it was necessary to sample the full range of inputs.  We
conducted sampling and determined pollutant loads for 5 tributaries, 19 point  sources
(including 6 municipal wastewater treatment plants and 13 pulp or paper mills), 3
atmospheric deposition stations, 12 storm sewers,  4 abandoned dumps or landfills, 60
groundwater locations, and 375 river sediment locations.  Sampling was  conducted to
characterize the mass of  pollutants in the open Bay at 27 water column sites and at 167
bay sediment sites.  The  water column sites were  sampled 8 times during 1988-90.
Sedimentation rates and  volatilization from the water column were characterized by
special rate  experiments.   Internal  recycling and storage studies were conducted for
water column interactions with sediments in Green Bay.  Biota studies were conducted
throughout the project by collecting phytoplankton and zooplankton at each of the
water column sites.  More than 3500 fish were combined into more than 685 samples
covering various food chains including forage base and several age classes of 3 top
predators.  The final Study Plan for  the project was adopted in April 1989 (4).

    All of these data were available for modeling  applications.  Existing hydrodynamic
models, water/sediment models, and food chain models were refined and calibrated. As
illustrated in Figure 3, the physical and chemical models were coupled with the food
chain models to estimate the pollutant body burden in target fish species, mainly carp,
brown  trout, and walleye.  We constructed  models to provide calibration, verification,
and prediction  of the conditions in the water,  sediment, and the biota under varying
combinations of regulatory and remedial action plans.  Both short-term  and long-term
predictive capability were developed.

    The sampling  efforts were particularly  successful and have resulted  in an
unparalleled database available for both current and future researchers.  Laboratory
quality was assured through Quality Assurance/Quality Control (QA/QC) protocols.
The ability of several laboratories to process samples on the specified schedules was an
issue that required continual  oversight by project managers.  The quality of laboratory
data along with the ability to provide low-level analyses in a timely way were issues that
required several modifications in the project scope and time schedules.  We
accomplished these modifications through a project management structure that involved
all of the project participants. The resulting data  could support far broader research in
the future as the full scope and detail of this database are made available to potential
investigators beyond those already included in the  project.

    In addition to the development and testing of the modeling component, project
managers conducted important and necessary activities for the management of the
project at both the technical  and administrative levels. Among these activities was  the
identification of the appropriate types of public involvement to  assure that the high
level of public  interest was not threatened  by a lack of understanding of the project.

                                         16

-------
 o
 Q.
 CD
 O
 C
 O
o


 CD

BE

 X
 o
D
 0)
 0
 L_
o
                                                                  0
                                                                 x
                                                                 o
 CD
Q_
 CD
Q


 E
 D
 CD

 CO
 Q_
                                                                       CO
                                                                       O
                                                                       *x
                                                                       o
                                                                             —
                                                                             "^
                                                                             T3
                                                                              O
                                                                              E

                                                                              C
                                                                              O
                                                                              O
                                                                              C
                                                                              V)

                                                                              D

                                                                              E

                                                                              L.
                                                                              0)
                                                                              X
                                                                              O
                                                                               D
                                                                              OQ

                                                                               C
                                                                               0)
                                                                               0)
                                                                               l_
                                                                              O
                                                                             K)

                                                                              0)
                                                                              L_
                                                                              3
                                                                              O)
                                       17

-------
                              RESULTS TO DATE
    Because all modeling will not be completed until late 1992, this paper can present
only preliminary results.  Even the preliminary results are very useful  for management
purposes.  Some of the earliest and most useful information is derived simply from the
mapping and mass estimates for contaminants in the sediments.

PCB RESULTS

    PCBs  have been measured, mapped, and modeled in Green Bay  and the Lower
Fox River. We determined the depth to which PCBs are found in the Green Bay and
Fox River sediments from the analyses of sediment cores.  From these data it is
possible to calculate the total mass of PCBs in the sediments.  Preliminary calculations
using these sediment cores indicate that the mass of PCBs in the system could be as
high as 63,000 kg.  Of this total, about 14,000 kg or 22% is in the bay (A. Andren,
personal communication, 1992), with the remaining 78% estimated to be in the Fox
River system (49,000 kg). The greatest portion of this mass of PCBs  resides below the
DePere Dam (45,000 kg) in the estuary-like, lowest reach of the Fox  River.  The
surveys and inventories have identified the proportions of the PCBs that are coming
from the various sources and moving into various compartments for the Lower Fox
River.  These initial mass balance estimates indicate that sediments are a substantial
source (more than 98%) of the annual PCB load to the  river above DePere (Figure 4),
and that the river is the major source (67%) to the Bay  itself, with point sources
comprising no more than 3 to 11% of the load to the Bay depending on the
assumptions made in using detection or less-than-detection values to compute point
source loads (5).  In addition, the early analyses  seem to indicate that the waters of
Green Bay may be a source of PCBs to the atmosphere (especially for lower-
chlorinated congeners) under many conditions, rather than serving as  a sink, as might
have been anticipated.  The volatilization of PCBs from  the water column probably
contributes more mass to the atmosphere than does atmospheric wet  or dry deposition
to the Bay.  More complete project results are required  before this conclusion can be
firmly or unconditionally validated.

    As shown in Figures 5-7, the bottom sediment PCB  flux to the water column for
model segments from Neenah to DePere is determined by flow characteristics (quantity
and channel configuration), sediment PCB content, and to a  lesser degree by total
suspended solids.  This later factor is illustrated by the several ordersof magnitude
difference in the summer and winter daily sediment fluxes.  In the summer,  the
suspended solids loads are increased because of algae and resuspension of bottom
sediments. These water column-sediment interactions show how the various deposits of
PCBs in the bottom sediments act as either sources or sinks  under the typical range of

                                       18

-------
 OS
Q

 0)
 0)
Q_
 0
Q

 £
 co
 0
  e
 x
 o
LL
t   u.
I    £
O)
CO
 °-   «
3   I
       CM
             CD
             •+-*
             03
            CD
            O
            0.
              0)
              o
              i_
              13
              O
              CO
              CO
                                     
L
                                     19

-------



~5
-A
\ .
l '
m
o
Q_

H_
C
CD
E
T5
(D
00
E
O
"o
DQ
i
^_
CD
^>
fV
LJ_

X
0
1 .
1 '











s*~ "^^
>^
1
i—
D
C
|
—
CD
s^_
Q.
L_
CD
C
$:
**-"*^










i ^-^" 	
L-— '
CD

CD 1
I
L_L_
CD 1
Qt ...

|
p-
L_



D
C
D
D \, |
V
C


- r —
CD 	
Q_ 1
^_ [Z

1
1 " '
_j_ m
D \
(D ^Sv
CD \
II IZ 1

—


—






-
-



-


_








CM

m
CM

~CM

,_ C
"CM O
_[_._..
^> 0



^.
o
L-
o>
C
£
L.
0)
-^ O ^-N C
O ^ =5
r*^ at* c.
^0
>
-^ o: ^
E E
o ~°

-« c § 1
S \ £
(-00
_ E VCM -Q
^ CO
CD
00
-en
Q) L
~rj
~^ o


-m


-ro


~'p~
E
-, O
s—
X
u

H —
O
Q_

in

3
tCT
^

m
                         CM
                            D  >-
x
o
                                                                     o
                                                                     CM
                          20

-------
 X
DO
O
Q_


 C
 

o:

 x
 o
                                                                       V)
                                                                       *
                                                                       *
                                                                             a)
                                                                             E
                                                                             E
                                                                             0>
                                                                             T5
                                                                             C
                                                                             C
                                                                             0)

                                                                             E

                                                                             TJ
                                                                             0)
                                                                              0

                                                                             E
                                                                             o
                                                                             E
                                                                             o
                                                                             i_
                                                                             M—

                                                                             X
                                                                             CD
                                                                             O
                                                                             Q_


                                                                             CD

                                                                             0)
                                                                             L.
                                                                             3
                                                                             O)
                               a)
                                         o >-
x
o
H-
O)
o
CN
                                       21

-------
 X
 IS
00
O
Q_
 Q)
 (D
      O
      c
      (D
     C
     (D
E^
2   ^
      D)
 x
 o
 Q)

 CD

 0)
Q

                                              c
                                              O
                      Q_

                      <
                                                                     r--
                                                                     
                                     (N
                                                                     to
                                                                     
                                                                     X
                                                                     o
                                                                                            o
                                                                                            CN
                                              22

-------
flow and hydrodynamic conditions. The models further indicate that these sources and
sinks can change their behavior significantly under other flow conditions (Figure 7).

    We used depth and areal extent estimates for the Fox River to quantify and map
the distribution of PCBs in the sediments.  The Little Lake Butte  des Morts river
reach is shown in Figure 8.  This reach is the location of some logical first targets of
opportunity for remediation of the contaminated sediments.  Specifically Deposit A, the
most upstream-known PCB deposit, accounts for an estimated 18.4% of the PCB mass
in the bottom sediments above the DePere Dam. This deposit provides an opportunity
for PCB remediation by possible removal or other means of deactivating this  important
source.

    At the lower end of the river (below the DePere Dam) the PCB transport model
indicates that a serious threat exists if the sediments were to be flushed  from their
current  location under high flow conditions. Field transport of PCB contaminants
(currently stored in the river sediment) into the open waters of Green Bay would
virtually quadruple the amount of PCB available to the biota of the Bay. This
increased load would constitute a major increase in PCB  availability in the waters of
Lake Michigan, which ultimately receive the outflow from Green Bay.  The review of
historic  events and dating of the sediments  indicate that such a flow event may have
occurred as recently as 1960, and remains quite possible in  the future.  Figure 9
identifies the major flow events in the period from 1960 to  1990 and correlates with the
sediment stratigraphy data for the major zones of deposition in the lower Fox River.

    The levels of accuracy found in these studies and models have interacted very
effectively in the early results. For example, the sediment survey data are sufficiently
precise and the models sufficiently accurate to  correlate well with the seiche effects on
the lower end of the river. The simulated hydrodynamics indicate  sediment deposition
along the northern shoreline and actually upstream (because of the seiche effects) of
historic  discharges of PCBs.  This condition was confirmed by field samples.  In
addition, the presence and the congeners of PCBs in the  sediment  stratigraphy correlate
well with the general knowledge of the historic wastewater discharge loads.

    The PCB transport model has already been used to identify the effects of
alternative  remedial measures especially on Deposit A,  but  also for others.  As
decisions are made on Deposit A, the effects and possible remedial actions at the
downstream deposits can be considered.  These modeling scenarios will be especially
useful in understanding the role of major events  involving high flows, seiche effects,
historic periods of high rates of PCB discharge, and evaluating the  effects of potential
remedial measures.

    Management options that have been analyzed to date for Deposit A are
summarized in Table 2.  These options demonstrate the value of the models used in
decision making.  The remediation of major sediment deposits may include inactivation
                                        23

-------
                   in
                   cu
                  -p
 in
Ld

 in
 in
 d
    CD CD CD CD  CD O) CD   CD
                                                                       L

                                                                       (U
                                in co
                                ^ CO
                       in
                       cu
                                              cr-
                                         -p
                                          I   cS
                         CQ
                               Q  LJ
    -P -P -P  -P  -P -P  QJ

PQ  in "in "in  in  "in "in n
O  o o o  o  o o
CL  CL a a a a CLID
    Cu QJ QJ  Cu  Cu QJ d
    ( ) f  ) Q  Q  (H) Q Q_
                                              o
                                              PQ
                                                        75
         (2

,°  CD  CU

\  U  ^


 d -5  QJ
PQ  C5 (X)
    PQ

 C     21
 Cu  U1 O
 cu  in m

 L  d

UD
 in
-p

 In
 o
 a
 Cu
o
                                                                                                                    0.01
                                                                                                                    < <"
                                                                                                                    jrCP;
 in
 QJ
 CU
-P
 QJ
                                                                                                              L.
                                                                                                              (LI
                                                                                                                                X

                                                                                                                                O
 cu

 o
_j

 QJ

-p


 o

 c
 o
-p

 o
 a

 d

 c

 in
-p
In
 o
 a
 cu
73

m
u
CL


 o
                                                                                                                                in
                                                                                                                                c
                                                                                                                                O
                                                                                                                                u
                                                                                                                                o
                                                                                                                               00

                                                                                                                                Qj
                                                                                                                                L
                                                                                                                                3
                                                                                                                                CD
                                                        24

-------
 (D
 >
ir
 D
 E
CO
x   °
o
_co
Q
      •<*-
      CM
                 CM
                 CM
                         O
                         CM
                              CO
                                             CM
                                                         OO
                                                                             CM
                                                                                      oo
                                                                                       I
                                                                                          CM
                                                                                          CO

                                                                                           I
                                                                                          .0
                                                                                           (D
                                                                                         00
                                                                                          I
                                                                                         >
                                                                                         o
                                                                                          I
                                                                                         >v
                                                                                         D
                                                                                          r--
                                                                                           I
                                                                                           I
                                                                                           L_
                                                                                           o
                                                                                         00
                                                                                       I
                                                                                       Q.
                                                                                       


-------
or isolation as well as removal.  Under the current schedule, remedial projects for
Deposit A may be initiated as early as the Fall of 1993.

        TABLE 2.  OPTIONS SELECTED FOR FURTHER CONSIDERATION FOR THE
          REMEDIATION OF DEPOSIT A IN LITTLE LAKE BUTTE DBS MORTS
Option
Intake Confined Disposal Facility
Upland Temporary Cell
Upland Confined Disposal Facility
Pretreatment With On-site Thermal Destruction
Fretreatment using P.H. Glatfelter Paper Company Combustor
Estimated Cost
($1,000,000)
3.5
7.2 - 13.3
2.7 - 10.2
15.2 - 17.3
4.9 - 6.5
Estimated Time
(months)
28
32-36
29-33
9
14
    The linkage of PCB transport models to food chain models is of special importance
to public health protection and general public  interest and support.  For these
mechanisms it is possible to predict fish contamination levels and impacts on other
aquatic organisms.

    Modification or control of sediment loads  from the tributaries must also be
considered as a way to change the sediment transport characteristics of the river.
Changes in tributary loads could result in increased deposition or scour at various sites.
Finally, the general management of flow rates  could provide an important tool.  The
river is highly managed by the presence of 14  dams and the inflow to the river is
subject to management of the storage capacity of Lake Winnebago. An evaluation of
the feasibility for control of peak flows to avoid scour and resuspension of PCB-laden
sediments will be  made.  In addition, perhaps  other flow management regimes could
result in the transport of selected sediment deposits from some of the less manageable
sites into some of the most manageable deposits—those deposits that  are most easily
subject to dredging, inactivation, or other sediment management techniques.

    The PCB transport model provides useful predictions in characterizing the changes
in the downstream load after  reductions in individual deposits.  Deposits A-F  (Figure 8)
in Little Lake  Butte des Morts account for approximately 50% of the PCB mass in the
bottom sediments above the DePere Dam.  The effect of the removal of all of these
contaminated sediment deposits is shown in Figure  10.

    In the larger  context, we will conduct simulations on the entire Fox River/Green
Bay system to evaluate management options from no action to extreme conditions such
as high flow events under catastrophic storm events. Table 3 identifies some  of the
management scenarios being considered.
                                        26

-------
CO
o
a.

01
 0
                   Q.
                   Q.
          o
          CO
o
o
CD
                               ca
                               c
O

LO
o
o
CO
o
CM
o
                                                     CM
                                                     LO

                                                     C\J
                                                     CO

                                                     C\J
                                      O)
                                                     LO
                                                  — CO
O
                                                             "O  i-

                                                             CD  0)
                                              TD

                                              CD

                                              E

                                              CD
                                              Q

                                              CD
                                                            I
                                              CO



                                              CO
                                                             C\J
                                            •O
                                            C
                                            Cti

                                            a) 2E

                                            °g
                                            (1) _l
                                            A _J

                                            to c
                                            c —
                                            o eo


                                            2 I

                                            c .i
                                            0) -o
                                            O Q)

                                            § M

                                            0 o
                                            m r-
                                                             CD   -O

                                                             00 .>  


                                            o>
                                 27

-------
     TABLE 3. MANAGEMENT SCENARIOS FOR WHICH THE GREEN BAY/FOX RIVER
                        MASS BALANCE MODEL WILL BE RUN

  Scenario                       Description

    1           No action run for a 25-year prediction period.

    2           Major event loading from a maximum scour event similar to 1960.

    3           Remediation of all areas above DePere that can be practicably removed.

    4           Extension of Scenario 3 with all areas below DePere Dam remediated.

    5           A 25-year projection for the Bay under 25, 50, and 75% reductions in the Fox River PCB loadings
                upstream of the DePere Dam.

    6           Peak flow reduction for the Fox River by control of the Lake Winnebago outlet dams.

    7           A 25-year projection on the impact of phosphorus load reductions due to nonpoint source controls.
DIELDRIN RESULTS

    The development of a mass balance for Dieldrin was hampered by analytical
problems and by the low levels of Dieldrin  that were found in the study area. At the
present, the development of mass balance models for Dieldrin are not anticipated.

CADMIUM AND LEAD RESULTS

    Because  of the low concentrations and  the lack of a concentration gradient, a mass
balance for cadmium could not be accurately developed.  Without accurate quantitative
data on the presence and mass of cadmium transported in the system, it is not possible
to develop a  useful model.  Accordingly, the study of cadmium was discontinued until
some future time when more sensitive analytical methods are available.

    Time constraints and the difficulty in providing low-level lead concentration
analyses have delayed the development of a mass balance model for lead.  However, as
would be expected, the Green Bay urban area is the major source of lead to the Fox
River/Green Bay system.  As lead additives are eliminated from gasoline, lead loading
from the urban area should decrease.  The  environmental effects of the current loads is
unknown at this time.

QUALITY ASSURANCE/QUALITY CONTROL CONSIDERATIONS

    Any report on this project would be incomplete without at least a brief note
concerning the management of QA/QC and the critical importance of adequate and
accurate laboratory facilities.  This project involved the use of 6 different laboratories
for the  analysis of PCB congeners. At the  beginning of the project there were concerns
that laboratory capacity and quality were inadequate for such an intensive project.
These problems were overcome by extensive commitment of existing laboratory capacity
                                          28

-------
and the development of additional  capacity outside of that which  could be  provided by
established laboratories.  Timely and accurate data were made available,  although often
requiring major project management efforts several times  during the  project period.  In
addition, the QA/QC protocols developed for this  project  were essential for the
credibility of the laboratory data critical to the technical success of the project (6).
Table 4 summarizes  the QA/QC approach used.  The appointment of  a Quality
Assurance Coordinator and a Data Manager  were extremely beneficial to the
maintenance of data credibility.

             TABLE 4.  COMPONENTS OF THE GREEN BAY/FOX RIVER QUALITY
                              ASSURANCE/QUALITY CONTROL PLAN
    Component

Field Blanks



Duplicate Samples

Duplicate Analyses


Procedural Blank



Matrix Spike




LOQ Verification
                   Comments
Surrogate Standard
Internal Standard
Comparability
Data Assurance/Management
Sampling container, material, solvent, or apparatus is taken into the
field and handled in the same manner as it would be for actual
sample collection. One field blank for every 9 field samples.

One duplicate sample is collected for every 9 samples.

A single sample is split and both halves are analyzed separately.  One
duplicate analyses is performed for every 9 samples.

Solvent and extraction apparatus are subjected  to the same analytical
procedures as samples. A procedural  blank is included with every
sample set.

A known amount of the performance standard  is added to a sample
matrix previously shown to be uncontaminated at a concentration
similar to sample concentrations.  Matrix spikes are included 10% of
the time.

A sample matrix previously shown  to be uncontaminated is spiked
with the performance standard at a concentration 2x-5x the target
LOQ of the smallest peak in the performance standard. This
QA/QC sample is included with every 20 analyses of samples
expected to have low concentrations.

A known quantity of a compound that is similar to the analyte of
interest, but not found in the environment is added to all samples
and blanks at  the beginning of the analytical procedure to monitor
recoveries of the analyte of interest.

A compound similar to the analyte of interest,  but not found in the
environment is added to all samples, blanks and standards just prior
to instrumental analysis to quantitate the analytes of interest.

All laboratories participating in data generation must analyze a series
of blindly coded  QA samples supplied by the QA Coordinator. The
QA requirements of the comparison must be met prior to any
sample analyses by the laboratory.

All data sets must be submitted to the QA Coordinator along with
all supporting QA/QC documentation. The QA Coordinator must
approve the quality of the data set before it can be released for use
by the Data Manager.
                                                   29

-------
OTHER SIGNIFICANT RESULTS

    Additional results anticipated from this effort are summarized below.

  A. Evaluation of the level of detail required for modeling to support management
     decisions envisioned, and a basis for  evaluating the management value and costs
     of less intensive monitoring and modeling efforts.

  B. Contribution of new scientific methods that have been developed during the  study
     and as part of the study (e.g., the study served as an active field test for the
     design and evaluation of sampling techniques and analytical methodologies for low
     concentrations of PCBs and metals).

  C. Development of shared insights among decision makers, politicians and scientists
     concerning the  alternative and selective management solutions.  This linkage is
     illustrated by the multi-dimensional video-tape (of modeled sediment transport)
     produced by Dr. Wilbert Lick at the  University of California-Santa Barbara.

  D. Strengthening and increased visibility of the Remedial Action Plan for Green Bay
     and the Fox River, with resultant public support for financial and regulatory
     actions necessary.

  E. Contribution to a strategy for sediment cleanup and specific engineering design
     for remedial measures - and thus development of a prototype for cleanups of the
     areas of concern around the Great Lakes.

  F. Development of other scientific results as this extensive and unparalleled data set
     is reviewed and known by other researchers.

            IMPLICATIONS FOR WATER QUALITY MANAGEMENT
    Many important lessons have been learned.  Only a few can be identified in this
brief paper but the following are some of the earliest noted.

  A. Development of a detailed, documented project study design is essential, including
     pre-planning field, laboratory and modeling needs to meet the desired project
     objectives. Even a dress-rehearsal sampling cruise contributed to problem
     prevention by early identification of problems and solutions, which otherwise could
     have jeopardized the coordination of the overall project field  sample collection.

  B. Quality assurance and  quality control for both laboratory and field operations are
     critical to the success of a mass  balance project of this type.  Our experience
     indicates that without such an effort, the data would not provide comparability
     and consistency in the database.  It is essential that we assure compatible

                                        30

-------
   sampling and laboratory analyses of conventional parameters -- especially
   important for total suspended sediment/solids.

C. Database management is critical if diverse and physically separated analysts and
   simulation modelers  are to work together with confidence and full ability to
   communicate.

D. Sediment is an active and critically important feature of such contaminated
   systems. The sediment reflects the results of pollutant contributions from pulp
   and paper mills, publicly owned treatment works, combined  sewer overflows, rural
   stormwater runoff from agriculture, urban stormwater runoff from acres of active
   daily use, as well as  raw materials and waste storage areas.

E. Sediment dynamics are widely variable and can best be understood with detailed
   studies.

F. Sediment modeling is invaluable to decision making. This point is critical for
   administrators,  managers, and politicians to understand.  The time and cost of
   conducting such sediment studies and developing modeling tools would be a
   difficult reality  for decision makers to accept.  This result, however, is a clear
   consensus of the experienced staff who have participated in  this project.
   Wisconsin Department of Natural Resources' staff have been seasoned by
   25 years of pollution abatement programs, and are veterans  of major successes in
   abatement of both conventional and toxic pollutants from point sources.  This
   same staff has  pioneered the modeling and abatement of nonpoint source
   pollution since  1978.  These staff members have already  found the sediment
   modeling tools  to be of greater use than were the  conventional water column
   pollutant modeling tools. Simulations of dissolved oxygen and biochemical oxygen
   demand were especially important tools in water quality successes achieved to
   date, but staff indicate that sediment modeling will be of even greater value and
   will lead to even greater savings and more intelligent decision  making.

G. The sediment,  hydrodynamic, water quality, and food chain models are all
   critically linked submodels of the overall environmental system  and represent the
   major components of an effective water quality management program.  As these
   models are linked, our insights and understanding  of the system have increased
   dramatically. The conceptual interaction of these subsystems has been made
   much more real to the analysts and decision makers than could have been
   otherwise possible.

H. The costs of mass balance studies are justified even at present  costs. At a total
   cost of over $10,000,000, the mass balance project is very expensive.  By contrast,
   however, the initial estimated costs of sediment cleanup for  the Fox River/Green
   Bay area are far higher. The Green Bay RAP of  1988 estimated that sediment
   clean up costs  could run as high as $640,459,000.   As  the first sediment cleanup

                                       31

-------
     experiences occur, it is evident that these mass balance models have great value
     for development of strategies, support of implementation, setting priorities, and
     for design of sampling and monitoring systems.  These tools are also of great
     value for analyzing and evaluating the results of the monitoring before, during,
     and after implementation of remedial actions.

                                 CONCLUSION
    State of the art advances may be expected to reduce the costs of mass balance
analyses in the future.  Studies such as these are costly. Based on this  project, it seems
there could be savings in the frequency, locations, precision, and even types of testing
undertaken. However, this project indicates that the study costs would  be justified,
even without the cost-savings that may come with future technologies.

                                  REFERENCES
1.   Fink, L.E., and P.L. Wise. 1988. Mass balance approach to water quality
    management in the Great Lakes Basin tributaries.  In: Protection of River Basins,
    Lakes,  and Estuaries, R.C. Ryans, ed.  Amer. Fisheries Soc., Bethesda, Maryland.
    pp. 21-32.

2.   Persson, L., V. Harris, C. Lukas, J.  Christie, H.J. Harris, L. Meyers, J. Sullivan, P.
    Allen, and R. Baba.  1988.  Lower  Green Bay Remedial Action Plan for the Lower
    Fox River and Lower Green Bay Area of Concern. Wis. Dept. of Natural
    Resources, Madison, Wisconsin. 319 pp.

3.   Ball, J.R., V.A. Harris, and D. Patterson.  1985. Lower Fox River, DePere to
    Green  Bay: Water Quality Standards Review.  Wis. Dept. of Natural Resources,
    Madison, Wisconsin.  69 pp.

4.   U.S.  Environmental Protection Agency.  1989.  Green Bay/Fox River Mass Balance
    Study Plan: A Strategy for Tracking Toxics in the Bay of Green Bay, Lake
    Michigan. Great Lakes National Program Office, Chicago, Illinois.  52 pp.

5.   Bierman, V.J., J.V. DePinto, T.C. Young, P.W.  Rodgers, and S.C. Martin.  1992.
    Development and Validation of an  Integrated Exposure Model for Toxic Chemicals
    in Green Bay, Lake Michigan - First Draft Final Report.  174 pp.

6.   Swackhamer, D.L.  1988. Quality Assurance Plan  Green Bay Mass Balance Study.
    U.S.  Environmental Protection Agency - Great  Lakes  National Program Office,
    Chicago, Illinois. 22pp.
                                        32

-------
               PROBLEMS RELATED TO MONITORING AND MODELING
                  POLLUTION FROM AGRICULTURAL WATERSHEDS

                                          by

                     V. Z.  Kolpak,1 T. A.  Dillaha III,2 D. B. Beasley,3
                         G. Y. Platonova1 and A. S. Voronkin1
                                     ABSTRACT

       Pollution from agricultural activities has only recently been acknowledged as a
significant danger to waterbodies.  Agricultural nonpoint source pollution is undoubtedly a
major pollution source in the Ukraine and the other republics of the former Soviet Union. We
review methods of evaluating agricultural pollution problems and relate these to conditions in
the Ukraine.  Models are one good evaluation method, and ANSWERS (areal nonpoint
source watershed environment response simulation) is the most accurate of presently
available models.
                                   INTRODUCTION

       Scientists have been aware of agricultural pollution for several decades, but since the
pollution was generally out of sight,  it was not treated as a significant danger to waterbodies
until recently. As agricultural production intensity and use of mineral fertilizers and pesticides
increased, negative environmental consequences of modern agriculture became more
apparent.  These consequences of poor agricultural production practices include: loss of
fertile top soil, growth of ravines and gullies, sedimentation of waterbodies, loss of nutrients
from top soil, excessive loadings of nitrogen and phosphorus to waterbodies with resulting
eutrophication problems, and movement of pesticides and their metabolites to waterbodies
where they negatively impact aquatic organisms. In the United States, the Environmental
Protection Agency has identified agricultural nonpoint source pollution as a major source of
pollution in rivers, lakes, reservoirs,  and coastal waters (Novotny and Chesters 1989).
In the Ukraine and  other republics of the former Soviet Union,  agricultural nonpoint source
pollution is identified as a problem.  The objective of this paper is to discuss some of the
Ukrainian Scientific Center for Protection of Waters, Kharkov, Ukraine.
2Biological Systems Engineering Department, Virginia Polytechnic Institute and State
 University, Blacksburg, VA (USA).
3Biological and Agricultural Engineering Department, North Carolina State University, Raleigh,
 NC (USA).
                                         33

-------
experiences we have had in attempting to assess agricultural nonpoint source pollution
problems on experimental watersheds in the Ukraine.  Recommendations on use and
development of water quality planning models for nonpoint source pollution control are also
made.

       Influences on the mass of nutrient and pesticide loadings to waterbodies are both
natural and anthropogenic.  Important factors are: climatic and hydrogeological conditions;
hydraulic and physical properties of watershed soils; distance between source
areas and receiving waters; chemical application rates, formulation, and methods;
degradability of agricultural chemicals; etc.
                                    DISCUSSION

      Several methods are available for quantitative evaluation of soil loss from agricultural
areas.  Shwebs and Formovanie (1974) reported four: 1) mark evaluation, 2) comparative, 3)
empirical dependencies, and 4) hydromechanical. Additional methods include conceptual,
physical, and mathematical models (Kuchment et al. 1983). The empirical dependency and
hydromechanical methods are the most widely practiced. Their development is the result of
accumulation and summarization of great volumes of experimental observations. Examples
of these approaches include the Universal Soil Loss Equation (USLE; Gudzon 1974) and the
methods of Shwebs and Formovanie (1974), Surmach (1976), and Mirtshulava (1970).
These methods have been improved over time, but their application in practice is still limited
because they do not estimate sediment loadings to receiving waters.  Use of these methods
alone leads to significant overestimation of sediment loadings to receiving waters (Novotny
and Chesters 1989).  Empirical methods are inherently limited because their application
conforms to the database from which they were derived. Unlike conceptual and physically
based models, empirical models cannot describe important processes affecting pollutant
movement such as sediment detachment, transport, and deposition in upland areas and
channels. Neither can they deal with the transport and fate of sediment-bound, dissolved,
and colloidal pollutants in ephemeral channels, streams,  and receiving waters.

      Effective management of pollutant loss from agricultural areas requires complex,
physically based models that simulate the significant processes affecting sediment and
chemical loss from  agricultural catchments.  These models must also consider effects of the
spatial distribution of catchment parameters,  and are often termed distributed parameter
models.  Such  physically based, distributed parameter models require four major submodels:
1) a geometric model representing catchment topography and drainage patterns
(mathematical image  of the catchment area), 2) a hydraulic model to simulate water
movement through the drainage network of the catchment, 3) an erosion and sediment
transport model, and  4) a chemical transport and fate model (nitrogen, phosphorus, and
pesticides).

      Physically based models are much more demanding in terms of data requirements
and mathematical calculations than empirical models.  In fact, solution of the equations
involved usually requires computers.  Data requirements for physically based, distributed
parameter models were so demanding that, prior to development and use of geographic
information systems,  they were rarely used.   While these models are more demanding of the
                                        34

-------
user, the value of improved sediment and chemical loss predictions often offsets additional
costs of using them.

       To accurately simulate chemical transport and fate in and from catchments, one must
first successfully simulate erosion and sediment transport, which is in turn dependent on
successful simulation of surface runoff by a hydraulic model.  The hydraulic model is
dependent on the geometric model.  Let us first consider the geometric or mathematical
image  model that seeks to represent catchment topography.  Distortions or errors in the
geometric model will  introduce errors in the hydraulic model, which will then introduce errors
into the erosion and subsequent chemical transport models. These errors  will be magnified
as they pass from one model to the next. Catchment topography can be represented by a
variety of geometric models, as shown in Fig. 1.

       The simplest geometric model, kinematic cascade (Fig. 1a), represents the catchment
as a series of overland  flow planes of varying slope (Kibler and Woolhiser 1970).  This simple
model  can be expanded to include channels.  Simple geometrical models composed of such
planes and channels include the Wooding water catchment approach (Fig.  1b) and the
Cherry (1976) approach (Fig.  1c).  Kuchment et al. (1983) represented the catchment area  as
a system of triangular and quadrilateral shaped planes and channels  that represent
catchment topography and ephemeral and perennial drainageways (Fig.  1d).

       Kolpak and Lysenko (1982) developed a geometric model that represents catchment
topography as a collection of triangular plains or cells of varying sizes that approximate
catchment relief in the best quadratic approximation (Fig. 1e); channels are also represented.
Complex watersheds may require hundreds of triangular cells and channels to accurately
reflect  site topography,  while simple catchments may require only a single cell.  Flowpaths
from any plain to the catchment outlet are a function of cell slope direction  and are
determined automatically  by the model.  Machmeier and Larson (1968) developed a method
of discretizing watersheds into discrete subwatersheds, which in turn  are divided into
elementary slope areas as depicted in Fig. 1f.

       Beasley (1978) represented catchments as a system of square elements or overland
flow plains ranging from 1/3 to 4 ha in size (Fig. 1g).  Concentrated flow through catchment
drainageways is modeled  with rectangular cross-section channel elements of specified slope,
width, and hydraulic roughness.  The model requires that all surface runoff exit the watershed
from a  single outlet cell and that the channel network be continuous.  In an approach similar
to Beasley (1978), Rumyantseva et al. (1986) represented watersheds as systems of square
cells, each with an area of 10  km2 (Fig. 1h).  For each cell, the steepness and direction of
waterflow into the river system are determined.

       Other geometric model approaches have been presented, but those discussed above
are the most  commonly used and accepted.  In our opinion, the combination of the
spline-function method and the method of Kuchment et al. (1983) with full automation of
computations on a personal computer is the best approach to follow in geometric modeling in
the long term. This approach  could be used to identify hydraulically active  areas as
proposed by Machmeier and Larson (1968) and to determine  flowpaths from  upland source
areas to the watershed  outlet (Kolpak and Lysenko 1982).  Whatever the geometric
approach,  development of a practical agricultural nonpoint source planning  model requires
                                        35

-------
Fig. 1.  Approaches to the approximation of watershed relief: a) kinematic cascade (Kibler and
Woolhiser 1970) and b) Wooding water catchment area, c) Cherry (1976), d) Kuchment et al.
(1983), e) Kolpak and Lysenko (1982), f) Machmeier and Larson (1968), g) Beasley (1978), and
h) Rumyantseva et al. (1986).
                                           36

-------
                      /
                     /
                    /
                 \/
                 \/
                    ''-^1>
                    '  -A
                    i \iT\/
  \ f
    \
    \


w\
                    \
                     \
                      N
                       \/
                       /


                       /
                      /
                      /
                  \/
  /
                       \ >p^
                       \x



                        \
                    /
 i i  •

 XXX

it +  + •
       X

    -V
      1+
      '-!-,
  +x^

                \
                 \
                     y/
                     r£^

Fig. 1 (continued).
               37

-------
coupling a geometric model with models describing surface runoff; erosion; sediment
transport; and dissolved and sediment-bound nitrogen, phosphorus, and pesticides transport
from upland areas.  The creation of a user-friendly program interface to facilitate data input
and interpretation of model output is also essential.

       At present, we believe that ANSWERS (Beasley 1978) is the most suitable model
currently available for water quality planning in agricultural areas. The ANSWERS model was
developed in the U.S. in 1977 for modeling runoff and sediment loss for short periods (1-2
days) from upland watersheds with complex topography, soils, and landuse. Since then,
ANSWERS has been modified to simulate particle  size distribution of transported sediment
and nutrient transport.

       ANSWERS is a distributed parameter watershed model; this means it can consider
spatial effects of topography, soils, and landuse variations on watershed response. This
allows the model to  be used to study effects of changes in small areas of the watershed on
overall watershed response.  ANSWERS represents a watershed as a system of square
overland flow cells and a channel network. Each of these model elements is described in
terms of its physical and chemical characteristics.  Important parameters described for each
element include: 1) break  point rainfall data; 2) surface and subsurface soil parameters; 3)
channel parameters, if a channel is  present; 4) surface conditions; 5) element slope and flow
direction; 6)  agrochemical characteristics; and 7) landuse and management. Additional
information on the ANSWERS model and its data requirements is given in a companion
paper in these proceedings "ANSWERS-2000: A Model for Agricultural Pollution Control
Planning in the USA and the Ukraine" and elsewhere (Beasley and Muggins 1982, 1991).

       Although the ANSWERS  model has been tested and verified in the U.S., Europe, and
Australia, it has  never been tested for Ukrainian soil, climate, and agricultural conditions. The
present study, therefore, was undertaken to provide data  for testing the model under
Ukrainian conditions.

       In our opinion, there has  been insufficient attention to model testing and validation
issues in the countries of the CIS.  This has often resulted in models that poorly represent
the physical systems simulated.  For example, Golosov et al. (1990) used the USLE equation
as modified  by Larinov (1984), to quantify erosion from small to  medium scale catchments.
Largescale maps of soil loss from these areas were then  produced without model verification
or investigation of the sensitivity  of the model to input parameters.

       In the CIS (COST  1984) and the U.S. (Barch et al. 1989), procedures have been
proposed to improve and  assess the quality of model input data with respect to: 1) precision
of measurements-the quality that reflects the repeatability or closeness of repeated
measurements;  2) accuracy of measurements-reflecting the closeness of measurement
results to the true value of the parameter measured; 3) reproducibility (comparability)--the
closeness and level of comparability of measurements performed by different researchers at
different times and locations in measuring the same parameter;  4) representativeness-the
degree to which measured data  taken from a limited number of samples of the system
reflects true  system values; and  5) completeness-the  quantity of data obtained in the
process of performing investigations in comparison with the quality of data under ideal
conditions.
                                        38

-------
       Alternative agricultural pollution management scenarios can be evaluated through
either mathematical modeling or monitoring.  If modeling is used, limited monitoring may also
be required for model verification and calibration. This is particularly important in regions
where the models have not been previously used. Surface runoff monitoring on
representative catchment areas enables collection of data on the volume and quality of
surface runoff for various precipitation and landuse conditions that can be used for model
verification purposes.  In addition, these data may be necessary for model calibration of both
physically controlled and uncontrolled model relationships.  Once the model is calibrated and
successfully verified for a particular region, the model can be used to assess losses of
sediment and agricultural chemicals in surface runoff from smaller or larger  catchments with
similar climatic, soils, topographic, and management conditions without additional monitoring
studies. Development of mathematical  models includes creation of a comprehensive list of
parameters required by the model. Designing systems to  collect these data often results in
improved monitoring systems.

       Main stages in the design of an effective monitoring system are: detailed definition of
purposes and scope of the proposed monitoring system; definition of required accuracy of
measured parameters; selection of a monitoring site; collection of preliminary technical data
on the site; and development of protocols for observations, field sampling and
measurements, laboratory procedures, and data processing. Details of the application of this
step by  step approach in designing a watershed monitoring system are reported elsewhere
(D. B. Beasley, G. J. Platonova, T. A. Dillaha,  V. Z. Kolpak and Y. M.  Plis, unpublished ms.,
Biological and Agricultural Engineering Department, North  Carolina State University, Raleigh;
and G. J. Platonova and V. Z. Kolpak, unpublished report, Ukrainian Scientific Center for
Protection of Waters, Kharkov).   These  reports give special attention to criteria for monitoring
site selection and for the collection of input data to which the ANSWERS model is the most
sensitive.

      The monitoring system design program was tested during development and
implementation of a surface runoff monitoring  system on two experimental watersheds in the
forest-steppe zone of the Ukraine.  The specific watersheds were located on the Martovoe
State Farm in the Boikov Yar River watershed that drains to the Pechenezhskoye water
reservoir, a primary water supply reservoir for the Kharkov region.  The larger monitored
watershed had an area of 586 ha and the smaller one, 2.4 ha.

      As a result of field and office investigations, the  representativeness of the monitored
Martovoe catchments  for the Left Bank Forest-Steppe region of the Ukraine was determined.
Topographic, soil, and agrochemical surveys were conducted to develop geometric models of
the catchments and monitoring protocols for obtaining soil and  agrochemical information.

      Since commercial equipment was not available in the CIS for monitoring and sampling
surface  runoff, we were forced to design and  construct  required monitoring systems including:
flumes, water level recorders of the "Valdai" type, semiautomatic discrete water samplers,
and a 1-m diameter Coshocton wheel (Brakensiek et al. 1979)  with accumulation tank for flow
proportional sampling  of runoff.  All samplers,  accumulation tanks, and flumes were
constructed from stainless steel to minimize pesticide adsorption problems.  The monitoring
system has been in operation for 3 years and  has been essentially failure-free.
                                         39

-------
       The most complicated aspect of the monitoring system was the characterization of
soils in monitored watersheds.  A soil sampling program had to be developed that would
supply data required by the ANSWERS model, as well characterize temporal and spatial
variations in soil properties in the watersheds.  The  soil sampling protocol includes full
descriptions of methodologies for field determination of soil parameters; sampling and
laboratory procedures; and quality assurance procedures specifying the required number of
samples, replications, frequency, and spatial  distributions of samples (G. J. Platonova and V.
Z. Kolpak, unpublished report).

       The monitoring system has been adequate for collecting data required by ANSWERS,
with the exception of soil infiltration parameters. Infiltration parameters were collected with
double ring infiltrometers.  Differences in measured  infiltration have varied by orders of
magnitude with small temporal and spatial changes in sampling.  Alternative methods of
determining infiltration  parameters, such as adding runoff to the upper end of a slope at a
known rate and measuring the runoff rate at  intermediate points down the slope (Nazarov
1981, Shemyakin et al. 1990), should also be considered. These methods, however, will
probably result in variations in infiltration parameters equal to or larger than the double ring
method.

       A monitoring system must also accurately record landuse  and  management factors for
each field or farm management unit.  Required types of data include:  field identification; crop
species; tillage and other cultural management practices; plant growth development dates
and cover conditions; dates of all tillage and management activities; and dates, methods,  and
rates of agrochemical applications.

       Our experience has shown that the physically based input data required for the
ANSWERS model can be obtained on experimental watersheds in the Ukraine. The
ANSWERS User's Manual (Beasley and Muggins 1991) was found to  provide reasonable
approximations of soil, cover, and landuse parameters required by the model for conditions  in
the chernozem zone, European territory, of the former Soviet Union and the Ukraine.  Work
by Roshkovan (1988) and Shwebs and Formovanie (1974) was useful in estimating potential
interception, surface roughness as a function of tillage and soil cover, and erodibility of
Ukrainian soils. Applicability of the USLE to the European territory of the former Soviet Union
was not experimentally investigated  in the present study.  However, other scientists (Larinov
1984, Kiryukhina and Pacukevich 1989) have adapted the USLE  soil erodibility nomograph by
developing correlations between particle size analysis procedures used  in the U.S.  and in the
former Soviet Union.  We, therefore,  assume that USLE parameters required by the
ANSWERS model can be reasonably estimated for  Ukrainian conditions.
                                   CONCLUSIONS

       Further enhancement of ANSWERS is necessary before it can be used as a practical
planning model.  First of all, it is currently a research model and, as such, it is not very user
friendly. For ANSWERS to be useful for routine planning purposes, it will require a user-
friendly program  interface.  The creation of an interactive interface for data input and display
of output would facilitate model use, and extend use to many more users.  Additional work
must also  be done to identify sources of information required to determine  model parameters
                                        40

-------
related to soil nutrient dynamics (soil sorption-desorption kinetics, mineralization rates,
nitrification-denitrification rates, etc.).  Inclusion of routines for simulating pesticides fate and
transport are also needed. Finally, ANSWERS needs to be converted into a continuous
simulation model to better represent processes affecting long-term sediment and
agrochemical loss. Development of the user-friendly program interface is the key task for the
near term.  Additional and expanded monitoring studies are also needed to accumulate data
needed for model development, testing, and verification.
                                   REFERENCES

Barch, D. S., B. J. Mason, T. H. Starrs and K. W. Broun.  1989. Soil sampling quality
       assurance user's guide, EPA 600-89X046. U.S.  Environmental Protection Agency,
       Washington, DC.

Beasley, D.  1978.  ANSWERS: a mathematical  model  for simulating the effects of land use
       and management on water quality.  PhD  Thesis, Purdue Univ., West Lafayette, IN.

Beasley, D. B. and L. E. Muggins. 1982.  ANSWERS (areal nonpoint source watershed
       environment response simulation): user's manual,  EPA 905X9-82-001, U.S.
       Environmental Protection Agency, Washington, DC.

Beasley, D. B. and L. E. Muggins. 1991.  ANSWERS user's manual. Agricultural
       Engineering Department, Univ. Georgia, Coastal Plains Experiment Station,
       Athens.

Brakensiek, D. L., H. B. Ozborne and W. J. Rawls.  1979. Field manual for research in
       agricultural hydrology. Agriculture  handbook 224,  U.S. Department of Agriculture,
       Washington, DC.

Cherry, D.  1976.  An approach to the simplification of watershed  models for application
       purposes. PhD Thesis, Utah State Univ., Logan.

Golosov, V. N., I. V. Ostrova and N. N. Ivanova.  1990.  Proektirovaniye
       Pochvo-vodo-okhrannykh  Meropriyatiy v Stepnoy Zone.  Pages 26-28 in Melioraziya i
       Vodnoye Khozyasyvo 6.

COST 16263-79.  1984. Metrologia.  Terminy i Opredeleniya-M.  55pp.

Gudzon, N.  1974. Okhrana Pochv i Borba s.  Eroziyej.  M. Kolos.

Kibler, D. and D. Woolhiser.  1970. Cinematic cascade as a hydrologic model.  Hydrologic
       Paper 39, Colorado State Univ., Fort Collins.

Kiryukhina Z. P. and A.  B. Pacukevich.  1989.  Erodiruemost  Pochv Evropeyskoy Chasti
       Sovetskogo Soyuza.  Pages 50-55 in Vestnik MGU. ser. 17, Pochvovedeniye, t. I.
                                        41

-------
Kolpak, V. Z. and V. E. Lysenko. 1982. Kontrol Kachestva Vody Netochechnykh Istochnikov
      Zagreniya s Ispolzonvaniem Matematicheskogo Modelirovaniya Formovaniya
      Poverkhnosti Stoka. Pages 128-134 in Kontrol Kachestva Prirodnykh i Stochnykh vod.
      Kharkov, Ukraine.

Kuchment, L. S., V. N. Demidov and Y. G. Motovilov. 1983. Formovaniye  Rechnogo Stoka.
      M. Nauka.

Larinov. G. A.  1984.  Metodika Sredne: 1  melkomasshtabnogo  Kartografirovaniya Eroziyno
      Opasnykh Zemel-Aktualnye Voprosy Eroziovedeniya.  Pages 41-65 in M. Kolos.

Machmeier R. and C. Larson.  1968.  Runoff hydrographs for mathematical watershed
      models.  American Society of Civil Engineering, J. Hydraulics Div.,
      November: 1453-1474.

Mirtshulava, C. E.  1970.  Inzhenernyye Metody Rascheta i  Prognoza Vodnoy Erozii.  M.
      Kolos.

Nazarov, G. V.  1981.  Gidrologicheskaya Rol Pochvy. L. Nauka.

Novotny, B. and G. Chesters.  1989.  Delivery of sediment and pollutants from nonpoint
      sources: water quality perspective.  J. Soil and Water Conserv. 44:568-576.

Roshkovan, D. N.  1988. Kinamika Stroitelnogo Pokrova v Agrolandshaftakh Moldavii
      (Gidrologo-eroziynyee Aspekty).  Kishenyev, Shtiniza.

Rumyantseva, V. A., S. A. Kondratyev, N.  I. Kapotova and N. A. Livanovan. 1986.  Opyt
      Razrabotki i Primeneiya Matematicheskikh Modeley Bassejnov Malykh Rek.  I.
      Gidrometeoizdat.

Shemyakin, N. M., V. A. Belolipski and I. A. Golovchenko. 1990. Konturno-meliorativ noye
      Zemledeliye na Sklonakh. K. Urozhay.

Shwebs, G. I. and V. E. Formovanie.  1974.  Stoka Nanosov i ikh Ozenka.  L.
      Gidrometeoizdat.

Surmach, G. P. 1976. Vodnaya Eroziya i Borba s ney.  L.  Gidrometeoizdat.
                                       42

-------
                 ANSWERS-2000; A WATER QUALITY PLANNING MODEL
                          FOR AGRICULTURAL WATERSHEDS


                                      by

         Theo A. Dillaha, III1, Vladimir Z. Kolpak2,  Faycal Bouraouri1,
                  David B. Beasley3,  and Galina Y. Platonova2


                                   ABSTRACT

      This paper describes ANSWERS-2000, a water quality planning model under
joint development in the USA and the Commonwealth of Independent States (CIS).
ANSWERS-2000 is being developed from the ANSWERS model, an event-oriented,
deterministic, distributed parameter, water quality model developed for
agricultural nonpoint source pollution control planning in the USA.  The
current version of the ANSWERS model simulates surface runoff and sediment
yield as a function of particle size.  ANSWERS-2000 will expand the model to
include phosphorus and nitrogen transport  (dissolved and sediment-bound);
long-term, continuous simulation; a user-friendly interface to facilitate data
file preparation and control model execution and output; and possibly a
pesticide transport model.

      This work is part of Project 02.02.11 "River Basin Water Quality
Planning and Management" of the joint USA-USSR Agreement on Cooperation in the
Field of Environmental Protection.  In the USA, support is being provided by
the U.S. Environmental Protection Agency,  Virginia Polytechnic Institute and
State University, and North Carolina State University.  In the CIS, the
Ukrainian Scientific Center for Protection of Waters (USCPW) is supporting
this work.
'Agricultural  Engineering Department,  Virginia  Polytechnic  Institute and State
 University, Blacksburg, VA 24061, USA

Ukrainian  Scientific  Center  for Protection  of  Waters,  6  Bakulina  Street,
 310888 Kharkov, Ukrain, CIS

'Biological and Agricultural  Engineering Department,  North Carolina  State
 University, Raleigh,  NC 27695, USA
                                      43

-------
                                 INTRODUCTION

      Water quality models are indispensable tools for studying processes that
influence water and chemical movement from  agricultural watersheds.   They are
also used by planners to assess the impacts of proposed best management practices
(BMPs) before  installing the BMPs.  Widely known and used  models,  developed
specifically for BMP evaluation  in agricultural watersheds, include CREAMS (1),
GLEAMS (2), AGNPS (3), and ANSWERS (4).   These models  use different approaches
to  simulate  the  complicated processes  of  hydrology,  erosion,  and  chemical
transport,  and all of them have limitations.

      CREAMS and GLEAMS, for example, are long-term, continuous models describing
sediment, nutrient, and  pesticide  transport,  but they are only  applicable to
field-sized areas with uniform soils and  cropping.  Their sediment and chemical
transport algorithms are  state-of-the-art, but the commonly used versions of the
models base their runoff calculations  on the curve number method  that poorly
represents runoff.  AGNPS  is an event-oriented,  watershed-scale, distributed-
parameter model  describing  runoff, sediment,  nutrient,  and COD  losses  from
agricultural watersheds.  AGNPS has a sophisticated user-friendly  interface that
facilitates data file creation and interpretation of model output,  but its runoff
calculations are  limited because they are  based  on the  curve  number method.
Nutrient transport algorithms are similar to those found in CREAMS.

      The current version of  the ANSWERS model has  significant limitations also.
It is event-oriented and cannot  predict long-term  average annual  losses; it has
nutrient transport versions,  but they have not been widely tested and verified;
it uses an infiltration equation whose parameters are  difficult to define; and
data file creation  is  tedious and  not  very user  friendly.   In  spite  of these
limitations,  ANSWERS offers an excellent  platform  from which a state-of-the-art
agricultural  nonpoint source  pollution  control  model   can  be  constructed.
Advantages of  ANSWERS,  compared with the  other  models,   include;  free public
access to the model  source  code,  a modular design  facilitating incorporation of
new model components, a state-of-the-art  surface runoff model,  and a data file
structure that is compatible with geographic  information  systems.   This paper
describes the  proposed changes  that are being made  to  the ANSWERS  model to
facilitate its use for agricultural nonpoint source pollution control planning
in the USA and the CIS

                              THE ANSWERS MODEL

      ANSWERS (Areal Nonpoint Source Watershed Response Simulation) was developed
in the late 1970s to address questions concerning the  water quality impacts of
agricultural activities (4).  The model was developed as part  of the Great Lakes
Program  in  the  midwestern USA.   Most  of  the  data  used  in  developing  and
validating the model is characteristic  of  the soils  and conditions  of  this
region.  The  earliest version of the model (5)  simulated surface water hydrology
only.  Beasley and  Huggins  (6)  added sediment  detachment  and transport to the
model.  The model was updated by Dillaha  and Beasley (7) to allow simulation of
sediment detachment  and  transport for different  soil particle  size  classes.
Dissolved and sediment-bound phosphorus (8)  and nitrogen (9) transport versions
of the model  have also been developed,  but they have not been documented well or
tested rigorously.
                                      44

-------
      ANSWERS La intended for use on ungaged watersheds and uses physically based
parameters to reduce or eliminate the need for calibration.  The ANSWERS model
also can simulate changes in management practices and is thus a valuable tool in
selecting among different management practice alternatives.   It also allows the
identification  of   subareas   within  watersheds   that  are  responsible  for
proportionally  more  off-site  pollution  than  the  watershed  as  a  whole.
Identification of these critical subareas allows water quality planners to focus
limited pollution control resources more cost effectively.

      The model divides watersheds into discrete square cells (up to 2000 cells
per  watershed)  with uniform  cropping, precipitation,  slope, soil  type,  and
management practices.  This allows the model  to  consider the  effects of spatial
differences  in watershed  properties on  watershed  response.   The  hydrology
component simulates  runoff, infiltration, interception,  surface  storage,  and
subsurface drainage.   Infiltration is computed using  Holtan's  equation (10).
Detachment of soil particles  by  raindrop  impact and overland flow is computed
using relationships based on studies by Meyer  and Wishmeier (11).  A modification
of Yalin's  equation  (12) is  used  to compute the transport  of  different size
sediment particles.

      One of the  major limitations of the current  version  of ANSWERS  is data
collection and data preparation.  About seven  parameters must be entered for each
cell.   This task can  be  tremendous for  large watersheds.   For very  large
watersheds, there is  a  dilemma:  should one select  large cell  sizes  (>1 hectare)
to reduce the amount  of input  data and computer  time required to run the model;
or should cell size be reduced to increase homogeneity within cells?

      The ANSWERS model has been  tested and verified on  watersheds in Illinois,
Indiana,  Iowa,  Ohio,  Oklahoma,   Texas,  Virginia,  Pennsylvania,   and  Canada.
However, researchers  have found that ANSWERS does not work as  well with European
and Australian soils  and cropping systems.  As a  result,  these researchers added
different  infiltration and erosion algorithms to  the model to  account  for
differences in soil characteristics.  These changes  have personalized the model
for the local conditions.

      Originally, the purpose  of  the joint research  described in this paper was
to determine the suitability of the ANSWERS model for nonpoint source pollution
control planning in agricultural areas of the Chernozem zone of the western CIS.
Early work  on  the model  in the  Ukraine identified  limitations of the  model.
Consequently development  of the new  version of the model,  ANSWERS-2000,  was
initiated to improve the model in the Ukraine and elsewhere.

                                 ANSWERS-2000

      ANSWERS-2000 is currently under development and a prototype version of the
model should  be  available  for release  and  testing in  August  1993.    Major
differences between ANSWERS-2000 and ANSWERS include: (a)  replacing the Holtan
infiltration model with the Green-Ampt infiltration model;  (b)  incorporating
dissolved and sediment-bound P and N transport into the model; (c)  use of leaf
area  index   data  to   represent   crop  growth  and   the  addition  of   an
evapotranspiration model to allow long-term simulation; and  (d)  development of
a user-friendly program shell to facilitate  data  file  creation  and to  control
program execution and output.   The proposed changes are summarized below.
                                      45

-------
INFILTRATION MODEL

      The  ANSWERS  model  currently  uses  a modified  version  of    Holtan's
infiltration equation (13):
                              FMAX = FC + A
                                                                          (1)
where FMAX - infiltration rate with surface inundated  (cm/hr); FC = steady state
infiltration rate (cm/hr); A = maximum infiltration rate in excess of FC (cm/hr);
TP = total porosity within the control depth (cm); PIV  « air remaining in control
depth before saturation (cm); and P = an empirical coefficient.

      The use  of this equation  is  limited because  of the  dependency of  the
infiltration rate on the control  zone depth (14).  This fictitious control zone
has no physical meaning and there are no uniform rules concerning the choice of
the  appropriate control  zone  depth.   For  example,   the ANSWERS user  manual
suggests a control zone depth of  one-half the A horizon or plow depth while the
FESHM model suggests a  control zone depth of the entire A horizon.  Furthermore,
this dependency  of  the infiltration rate upon  the control  zone  depth  defies
hydraulic principles (14).  This equation is difficult to use because A and P are
best determined by calibration.

      The Green-Ampt infiltration equation  (Green and Ampt,  1911) was chosen to
replace Holtan's equation for  three  major  reasons.  It is a physically  based
approach, it is computationally efficient, and its parameters can be determined
from readily available soil and vegetal cover  information.   The  Green-Ampt
equation can be expressed as

                             K -F -N. inll + JLI                          (2)
                                    - M. ln|l * -t
where K, = effective saturated hydraulic conductivity (cm/hr); F  =  cumulative
infiltration (cm), and N, = effective matrix potential (cm).   Infiltration rate
is determined by differentiating Eq.  2.

PHOSPHORUS  MODEL

Sediment-Bound Phosphorus Model

      The  sediment-bound P  transport  model  assumes  that  sediment-bound  P
transport is directly proportional to sediment transport.  The sediment transport
model  simulates  detachment,   transport,   and  deposition  of  sediment.  The
conservation of mass  of  P for each particle  size  class  and each cell  can be
expressed as

                                  »-'-                                 <3>
where Pj  = sediment-bound P inflow from surrounding cells and  soil  detachment
within the cell  (kg/s);  P0 = sediment-bound P outflow to adjacent cells  and P
loss due to deposition within the cell (kg/s);  P = sediment-bound P  in transit
(kg); and dt = model time step (s).


                                       46

-------
       Equation  3  is solved  at each time  step for each  particle size  class
separately.  The initial P content of the soil is distributed among the different
particle size classes according to the specific area of the particles, such that

                               PSSA = POSOILX10'>                           (4)
                                         SSAT

where PSSA = P content  of the soil (kg/m2); POSOIL = P content of  the original
soil (jug-P/g-soil) ; and SSAT = specific  surface  area of the original soil (mz/g-
soil).

      The concentration of P  for each particle class, PO,,  is

                                 PO, = SSA, PSSA                             (5)

where SSAj  is the specific surface area  for  particle  size class  i (m2).

      The basic  assumptions of  the sediment-bound P model are
      1.    The  sediment   transport  model  accurately  simulates   sediment
            transport .
      2.    Sediment-bound  P is  distributed between  the soil  particle  size
            classes in proportion to the  specific surface area of the individual
            particle size classes.
      3.    There  is  no redistribution  of  P between  different particle  size
            classes during or after  a storm.

Dissolved Phosphorus Model Development

      An extraction function  (16)  is used to describe the release  of  dissolved
P from surface soil to runoff:
                                     * ta * WS" * P
                                                .,,
where Pd = cumulative P desorbed (/jg-P/g-soil) ; ?„,= initial extractable  soil  P
level (^g-P/g-soil) ;  t = time of contact  (minutes);  WS  = water to soil  ratio
(cm'/g);  and K^,, a, P = empirical soil type specific constants.

      The water  to soil ratio is defined  as  the  ratio of the rainfall  volume
falling during the time increment and the mass  of suspended sediment. Phosphorus
is desorbed from the mixing zone or  effective depth of interaction (EDI),  which
is defined as the depth of soil at  the soil surface that  interacts with surface
runoff.  The amount of P desorbed during the time increment can be expressed as
                                  .
                                dt   EDI  DX2  pfc
                                                                           (7)
where C = concentration of dissolved P in runoff  (/ug/L); EDI = effective depth
of interaction (cm); I = rainfall volume during time  increment  (L);  DX2 - area
of the element (cm2);  and pb = soil bulk density (g/cm3) .

      The concentration of dissolved P can be derived from Egs. 6 and 7 as
      If runoff occurs,  then the loss of dissolved P is the product of the runoff

                                      47

-------
                             a Km f - ' WS" Pm EDI DX2 pt                      (8)
at the end of the time increment and the average concentration of dissolved P in
the cell.  The  average  concentration  of  dissolved P in the cell is the volume
weighted P concentration  due  to dissolved P inflow from neighboring cells and
storage and extraction  from the given cell.

      At the  end of each  storm, a  P balance is computed and an average soil P
concentration is determined.  This average P concentration is used as the initial
soil P concentration at the beginning of the next storm event.

      The assumptions made for the dissolved P model are
      1.    Precipitation  and  living  and  decomposing  plant  material  are
            negligible  sources of P.
      2.    P desorption is diffusion controlled.

Phosphorus Uptake by Plants

      Plants  absorb available forms  of inorganic dissolved  P.   The following
equation  (17),  developed  for describing  N uptake by plants, is  also  used to
describe P uptake:

                                   UP = AU J|                              (9)


where UP = plant uptake of nutrient by plants  (kg/ha); AU = available nutrient
concentration (kg/ha);  ET  = evapotranspiration  (cm); and SW =  average soil water
content (cm).

      Phosphorus uptake is  assumed to occur uniformly  throughout the rooting
depth.  If the root depth  is  larger than  the average depth to the wetting front
(depth to which infiltrated water has penetrated), nutrient uptake  is reduced to


                                  UPZ  = UP -±-                              (10)
                                           Rd

where UP, = reduced amount of P uptake  (kg/ha);  L = average depth to the wetting
front as predicted by the Green-Ampt  infiltration model (m); and Rj = rooting
depth (m).   If the root depth is  less than the  average depth  to the wetting
front, uptake  is distributed uniformly down to  a depth equal to L.  The nutrient
uptake by the plant model will be used between storms to estimate soil phot phorus
levels at the beginning of runoff events.
NITROGEN MODEL

      The  proposed  model simulates  the transport of  the following  forms of
nitrogen: dissolved  nitrate and ammonium, and sediment-bound ammonium and organic
N.  Nitrogen  processes  modeled  include  mineralization,  nitrification,  and
adsorption/desorption of ammonium.


                                       48

-------
Mineralization

      Mineralization, or conversion of organic N to nitrate between  storms,  is
simulated as  (18)

                           MIN = POTM — (1- B-"7* )                       (11)
                                      FC

where MIN = mineralized N  (kg/ha); TK = temperature coefficient; AWC =  average
volumetric  water content  (mm3/mm3);  At  =  time  increment  (days); PC  =  field
capacity (tm3/im3)} and POTM = potentially mineralizable N  in the soil  (kg/ha).

      The transformation of NH4+  to NO3' through nitrification between storms  is

                           DNI = N03 * (1 - e"*7"* (W-oflj                      (12)

where DNI = amount of denitrification (kg/ha); NO3 = nitrate in  the  surface zone
(kg/ha); DT = duration  of drainage  since  the  last  storm (days);  and DKT =
temperature adjusted rate  constant.

Sediment-Bound Nitrogen Model

      The sediment-bound  N model considers  ammonia  and  organic N separately.
Transport of each sediment-bound N form is simulated with the same equations used
by the  sediment-bound P model except that  the ammonia and organic  N are the
species being simulated.

Dissolved Nitrogen Model Development

Nitrate Transport—

      The nitrate transport model assumes that all nitrate in the effective depth
of interaction is available for transport. Thus nitrate transport from a cell  is
the product of the total amount of nitrate in a cell, times the volume  of  outflow
from the cell, divided by  the total volume of water in the cell.

Ammonium Transport—

      Dissolved  ammonium  transport  is more complex  because  of  the  dynamic
equilibrium  existing between the dissolved and attached  forms of  NH4+.  The
Freundlich  equilibrium isotherm  (19) is used to express the equilibrium  between
the two phases of NH«+

                                   St = K C1/"                               (13)
where  Se  =  concentration  of  ammonium  adsorbed  by  soil   (pg/g-soil);  Ce =
concentration of dissolved ammonium  (mg/L); and K and n = constants.

Plant Uptake of Nitrogen

      The  predominant form  of N  available  is  NO3~ because under  most soil
conditions NH4+-N is  rapidly transformed to  NO3"-N  (nitrification).   Ammonium
becomes the major  form of available N to plant under  conditions that are  not

                                      49

-------
favorable to the nitrification process  (20).   In  the proposed model, only the
uptake of  NO3"  is considered.  Plant  uptake of N is simulated with the same
equations used to model P uptake.

EVAPOTRANSPIRATION MODEL

      Ritchie's method  (21) will be used  to estimate evapotranspiration (ET).
Incorporating  ET is a necessary  step  in making ANSWERS-2000 a continuous model
because  ET depletes the soil moisture  that establishes infiltration  at  the
beginning of each storm.

      The Ritchie's approach was  selected because it uses the leaf area index as
a parameter.   This  approach allows ET to be simulated  as a function of plant
surface condition and cover.  Transpiration from plants and soil evaporation will
be simulated separately because parts  of watersheds may be vegetated while some
others may be bare.  The ANSWERS model  does not divide the soil into layers,  and
the only layer used is that  defined by the soil surface and the average wetting
front depth. Evapotranspiration will remove water from the zone defined by the
average wetting front depth.  The soil moisture will be assumed to be uniform to
the average wetting front depth.

      Evapotranspiration will be computed on a daily basis  between storms  for
each combination of  plant cover  and  soil type.   The  fraction of  potential ET
(FETP) transpired is determined by
                          FETP =   ETP30M (l - e-3-0651^)                     (14)


where RZ is the root zone depth (cm) and ETP is the potential ET predicted by the
Ritchie model.   If the water requirement for ET exceeds the amount  of water
available  (water above residual content), ET will be reduced accordingly.

USER-FRIENDLY INTERFACE

      A user-friendly interface is also being developed for ANSWERS-2000.  The
interface  or  shell has three  principal  functions:  (1)  input  and  editing of
required data; (2)  control of model execution; and  (3) control of model output.

      The shell will assist in creating and entering data required by the model.
It  will first  assist  in creating  the  required  basic  elemental  data  file
containing topographic,  soils,  landuse, and channel  information required by the
basic model that simulates surface runoff  and sediment transport.  The shell will
eventually include a link with a public domain geographic information system to
facilitate entry  and  manipulation of  elevation,  soils,  landuse, and channel
network data.  Help screens  will  be provided to guide the  user through  the
creation of the database.

      The shell will then request information on the parameters to be simulated,
whether the model  is to be run  for a single  event  or in  continuous  simulation
mode.  The shell will  then query the user for the data required for the type of
simulation desired. For  example,  if  only P  losses  are required  from  a single
design storm,  then  the shell will build the data file required  for a single event
simulation and for running the P model  only.
                                      50

-------
      Another portion of the shell will graphically display the watershed, after
the basic elemental data file has been created,  to enable the user to specify the
points in the watershed  at which output is desired.

      Finally, the  shell will  include a graphical  output section to assist in
interpretation of the model's output  and to identify cells with proportionally
higher sediment and nutrient losses.  The shell  will allow the user to adjust the
parameters of cells with high pollutant losses  to simulate  alternative BMPs and
then run the model  again.  This process can be repeated until a system of BMPs
is identified which results  in acceptable pollutant losses.  Model output will
be assessable by the geographic information system used in creating the original
watershed database.

                         MODEL TESTING AND VERIFICATION

      The model will initially be tested on four watersheds in the USA and the
Ukraine.  The two Ukrainian watersheds are agricultural and representative of the
Chernozem region of the western CIS.   The watersheds are located on the Martovoe
State Farm near Kharkov.  One watershed is  586  ha and the other is 2.4 ha. Both
field and vegetable crops are grown in the watersheds and cropping  rotations are
fairly complex.  These  watersheds  are described in another symposium paper by
Kolpak and Platonova.

      In  the  USA,  the model will  be tested on the Owl Run  and  Nomini  Creek
watersheds in Virginia.  These watersheds are predominately agricultural and have
been monitored since  1986.   The Owl  Run watershed  is  in the Piedmont region,
while  the  Nomini   Creek watershed  is  in the  Coastal  Plain.    Runoff  and
precipitation are monitored  continuously from these 1200-ha watersheds.

                            SUMMARY AND CONCLUSIONS

      ANSWERS-2000, a physically based,  distributed-parameter, continuous water
quality model is currently under development in the USA and CIS.  The model is
intended for evaluating the effects of agriculture and best management practices
on water quality.   The  model will  initially simulate surface runoff, sediment
transport, and dissolved and sediment-bound nitrogen and phosphorus losses.  The
model will be interfaced with a geographic  information system and  user friendly
shell to facilitate data file creation, model  execution and interpretation of
model output.  A preliminary version  of the model  is  scheduled  for release in
August 1993 for testing.
                                       51

-------
                                  REFERENCES

1.    Knisel, W. G.  1980.  CREAMS: A Field  Scale  Model  for Chemical,  Runoff,
      and Erosion  from Agricultural  Management Systems.   Conservation  Res.
      Report No. 26, USDA-SEA.  643 pp.

2.    Leonard,  R.A.,  W.6. Knisel  and  O.A.  Still.  1987. GLEAMS:  groundwater
      loading effects of  agricultural management systems. Transactions of the
      ASAE 30(5):1403-1418.

3.    Young, R. A., C. A.  Onstad,  D.  D.  Bosh, W. P. Anderson.   1987.   AGNPS,
      agricultural nonpoint source pollution model: A watershed analysis tool.
      U. S.  Department of Agriculture, ARS,  Conservation Research Report No. 35,
      Washington D.C.   80 pp.

4.    Beasley, D.B., L.F.  Huggins,  and E.J. Monke.  1980.  ANSWERS: A model for
      watershed planning.  Transactions of the ASAE 23(4):938-944.

5.    Huggins, L.F., and  E.J.  Monke.  1966.  The mathematical simulation of the
      hydrology of small  watersheds.  Technical Report No.  1.   Water Resources
      Research Center, Purdue University.  West Lafayette,  Indiana.  130 pp.

6.    Beasley,  D.B.,  and  L.F. Huggins.   1982.   ANSWERS user's manual.   EPA-
      905/9-82-001.  U.S. Environmental Protection Agency,  Region V.  Chicago,
      Illinois.  54 pp.

7.    Dillaha, T.A., and D.B.  Beasley.  1983.   Distributed parameter modeling of
      sediment movement and particle size  distributions.  Transactions of the
      ASAE 26(6):1766-1772,1777.

8.    Storm,  D.  E., T.  A.  Dillaha III,  S.  Mostaghimi,  and V. O.  Shanholtz.
      1988.   Modeling phosphorus transport in surface runoff.   Transactions of
      the ASAE  31:117-127.

9.    Dillaha, T.A., C.D.  Heatwole, M.R. Bennett, S. Mostaghimi, V.O. Shanholtz,
      and B.B. Ross. 1988.  Water quality modeling for nonpoint source pollution
      control planning: nutrient transport.  Report No.  SW-88-02.  Department of
      Agricultural  Engineering,  Virginia  Polytechnic  Institute  and  State
      University.  Blacksburg, Virginia.   117 pp.

10.   Holtan, H. N.  1961.  A concept  for  infiltration estimates in watershed
      engineering.  USDA-ARS Bulletin 41-51.  25 pp.

11.   Meyer, L. D. and W. H. Wishmeier.   1969.  Mathematical simulation of the
      processes  of  soil  erosion  by   water.     Transactions  of  the  ASAE
      12(6)754:758.

12.   Foster, G.R., J. D.  Ross, and L. J.  Lane.   1980.  The erosion submodel.
      In; CREAMS:  A Field Scale Model for Chemical, Runoff,  and  Erosion from
      Agricultural Management Systems,  Knisel, W. G. (ed.).   Conservation Res.
      Report No. 26, USDA-SEA. pp. 37-64.


13.   Overton, D. E. 1964.  Mathematical refinement of  an infiltration equation

                                      52

-------
      for watershed engineering.  USDA-ARS 41-99.  11 pp.

14.   Smith,  R.   E.     1976.    Approximation   for  infiltration  patterns.
      Transactions of the ASAE 19(3):505-509.

15.   Green, W. H., and G. Ampt.  1911.  Studies of soil physics, Part I.  The
      flow of air and water through soils.  Journal of Agricultural Science 4:1-
      24.

16.   Sharpley, A. N., L. R.  Ahuja,  and R. G. Nenzel.   1981.   The release of
      soil phosphorus to runoff  in relation  to the kinetics of desorption  J.
      Environ. Qual. 10:386-391.

17.   Frere, R. G., J.  D. Ross, and L. J. Lane.   1980.  The nutrient submodel.
      In: CREAMS:  A Field  Scale Model for Chemical, Runoff,  and Erosion from
      Agricultural Management  Systems.   W.  G.  Knisel,  ed.  Conservation Res.
      Report No. 26, USDA-SEA. pp. 65-87.

19.   Freundlich,  H.    1926.    Colloid  and  capillary chemistry.   Methven  &
      Company, London.

18.   Stanford. G., and S. J. Smith.  1972.  Nitrogen mineralization potential
      of soils.  Soil Sci. Soc. Proc.  36:465-472.

20.   Haynes, R.  J.,  Cameron. K. C.,  K. M.  Goh,  and  R. R.  Sherlock.   1986.
      Mineral nitrogen in the plant-soil system.   Academic Press, Inc.

21.   Ritchie, J.  T.  1972.   A model  for predicting evapotranspiration from a
      row crop with incomplete cover.  Water Resources Research 8(5):1204-1213.
                                      53

-------
     RELATIONSHIPS BETWEEN WATERBORNE TETRACHLOROGUAIACOL
               UPTAKE AND OXYGEN CONSUMPTION BY FISHES
                                         by

               J.F. Neuman1, R.V. Thurston1, C.J. Brauner2 and D.J. Randall2


                                    ABSTRACT

      The objective of the studies reported here was to determine whether there is a correlation
between oxygen consumption by fishes and bioconcentration of tetrachloroguaiacol (TCG).  Two
species of fishes were tested—largescale suckers (Catostomus macrocheilus) and rainbow trout.
Test fish were exposed to known TCG water concentration under controlled ventilation
conditions using a modified Brett-type respirometer where oxygen consumption could be
dynamically measured. Individual whole body TCG concentrations were determined by Ni63
electron capture gas chromatography. Lipid levels were also determined.
                                 INTRODUCTION
      The presence of chloroguaiacols and other chlorinated phenols in Kraft pulp mill bleach
effluent was reported as early as 1975 (Leach and Thakore, 1975) where 3,4,5-trichloroguaiacol
and 3,4,5,6-tetrachloroguaiacol (TCG) were isolated from effluents of paper manufacturing
facilities in northwestern Canada. Static toxicity tests on these compounds using juvenile rainbow
trout (Ocorhynchus mykiss) indicated that both compounds were toxic with 96-hour LC-50
'Fisheries Bioassay Laboratory, Montana State University, Bozeman, MT 59717, USA
Department of Zoology, University of British Columbia, Vancouver, BC, V6T 2A9, Canada
                                        54

-------
measuring at 0.32 mg/liter TCG.  Guaiacol (2-methoxyphenol) is a naturally occurring product
derived from lignin cleavage (Dence 1971) during paper production.  Once cleaved,
chlorination occurs during bleaching and eventually TCG and other chlorinated toxic
byproducts are discharged with the caustic effluent, including resin acids (Leach and Thakore
1973).

       Chlorinated guaiacols and phenols were also detected in pulp mill effluents located on
the shores of the Baltic Sea where accumulation of these components in rainbow trout were
measured by exposing test fish to caustic extraction effluent from milling operations (Landrer
1977).  Significant uptake of at least 3 different chlorinated phenols: 2,3,6-trichlorophenol,
trichloroguaiacol, and TCG was observed.

       Swedish investigators (Renberg 1980) measured bioaccumulation (BCF) and toxicity
(96-hour LC-50) of TCG to bleaks (Alburnus alburnus) native to the brackish seawater located
in the vicinity of Kraft pulp plants along the Baltic Sea and also observed toxic effects and
reproductive degeneration in Crustacea (Nitocra spinipes) exposed to TCG.

       Taste-impaired fish collected near Kraft pulp mill operations in Finland were found to
have elevated concentrations of chlorinated veratroles and anisoles including TCG (Paasivirta
1987).

       Bioconcentration factors (BCF) of TCG and other chlorinated phenols present in pulp
and paper mill effluent were measured in subadult (mean weight 140 g) rainbow trout (Niimi
1990).  These fish were exposed to TCG at waterborne concentrations of 1 /ig/liter .
Equilibrium was reached in 2 days with fish averaging 189 /xg/g (log BCF =  2.26). Log P
for TCG was also determined at 4.28.  Niimi stated that these kinetics are consistent with
residue levels reported in fish collected from waters receiving mill effluent where water borne
uptake appears to be the primary mode of accumulation. Dietary exposure was also evaluated
and it was found that TCG is poorly absorbed, and has a half-life of less than 2 days.

       The objective of the studies reported here was to determine if there is a correlation
between oxygen consumption by fishes and bioconcentration of TCG. Two species of fishes
were tested: largescale suckers (Catostomus macrocheilus) and rainbow trout.  Test fish were
exposed to known TCG water concentrations under controlled ventilation conditions using a
modified Brett-type respirometer (Gehrke et al. 1990) where oxygen consumption could be
dynamically measured. Individual whole body TCG concentrations were determine by  Ni63
electron capture gas chromatography.  Lipid levels were also determined.

                           METHODS AND MATERIALS

Fish exposure - respirometer protocol

       Largescale suckers (400 - 900 g) used in this study were wild fish collected from
Cultass Lake located in British Columbia, Canada.   Large (340 - 490 g) and small (4.6-5.8
g) rainbow trout used in these tests were procured from fish farms  in British Columbia. All

                                         55

-------
fish were held in aquaria prior to testing at the Department of Zoology, University of British
Columbia, Vancouver, British Columbia.  At the time of test, individual large fish or groups
of small fish were introduced into the respirometer and allowed to accumulate at the test water
temperature for at least 1 hour prior to dosing. During this time, the fish were swum at low
swimming speeds,  < 1 body length per second (Bl/s). All fish were exposed to the same
initial TCG water concentration of 200 /xg/liter by the addition of 26 mg TCG (in 67 ml
acetone) to the water at the start of the test.  Once toxicant was added, the system was sealed
and the water flow adjusted to a predetermined speed controlling ventilation.  Five minutes
after dose addition, flow was stopped and the system opened  for 2 -3 minutes so that water
samples could be collected for subsequent TCG analysis.  The system was again sealed, the
test continued, and oxygen depletion measurements initiated and continued throughout the
duration of the test. All fish were them swum for a 2-hour period, at the
end of which water samples were again taken for analysis. The fish were removed from the
respirometer, rinsed briefly with tap water and sacrificed. Each fish was weighed and
measured and then placed in a labelled plastic bag. All chemical analyses  on both fish and
water samples were performed at Fisheries Bioassay Laboratory, Montana State University,
Bozeman, Montana. After each test, samples were either shipped by overnight commercial air
carrier to Bozeman, packed with dry ice, or stored at -80°C until shipment. Once received in
Bozeman, samples were also stored on dry ice until analyses.

       All fish used in the study were exposed to TCG at two different ventilation rates. For
tests involving largescale suckers, five fish were individually placed in the respirometer and
exposed to TCG at a low ventilation rate where the water temperature was 9°C.  A second set
of five fish were subjected to the same regime, but a higher ventilation  rate was induced by
using a water temperature of 18°C. These two temperatures  were used to achieve different
oxygen consumption rates by largescale suckers rather than dissimilar swimming speeds
because these fish were unable to swim at speeds much above 1 Bl/s. For the tests with
rainbow trout, different ventilation rates were achieved using various swimming speeds.  Five
large rainbow trout were individually exposed to toxicant at a low ventilation rate with
swimming speeds of 1.3 -  1.4 Bl/s. Then  a second group of  five fish were tested at  a higher
ventilation rate using swimming speeds of 1.7 - 1.9 Bl/s.  In  all these tests the respirometer
water temperature was maintained at 7°C.  Two groups of small rainbow trout were tested;
one group (n=27) was swum at rates of 2.1 - 2.5 Bl/s, the second (n=26) from 3.2 - 3.9 Bl/s.
The respirometer water temperature during these tests was held at 5°C.

TCG analysis - Fish

       Fish preparation, surrogate addition, and extraction - Each large fish was partially
thawed, cut into several pieces and then ground using a conventional kitchen meat grinder.
The ground sample was further mixed to achieve as homogeneous a tissue sample as might be
possible. This homogenate was placed in a labelled plastic bag and stored on  dry ice or in a
freezer at -80 °C until analysis for TCG and lipid content. TCG analyses were performed in
duplicate on portions of the homogenate taken from  different parts of the sample, and on
different days.  At the time of extraction, weighed aliqouts (ca. 5 g) of homogenate  was placed
in square wide mouth extraction bottle. Small fish were weighed and then diced into small

                                         56

-------
pieces while still frozen.  These pieces were placed into a tared square-shaped extraction bottle
and the weight recorded.  This weight was used as the fish weight when calculating final TCG
concentrations. Each sample, either homogenate aliquot or diced small fish, was acidified by
the addition of 2.0 ml 18N sulfuric acid followed by thorough mixing.  Surrogate
(pentachlorophenol, 10 /xg) was then added to each sample followed by 80 ml extraction
solvent (1:1 acetone:methylene chloride).  Extraction was achieved by blending this mixture
for 30 seconds using a Polytron™ tissue homogenizer. The extract was decanted through
anhydrous sodium sulfate into a 250-ml volumetric flask. The residue extracted again, both
extracts were combined, and the volume adjusted to 250 ml with additional extraction solvent.

       Extract methylation and cleanup - Aliquots (5 ml) of extract were taken to near dryness
with the aid of a gentle stream of dry nitrogen and low heat.  The residue was immediately
methylated (^ 30 minutes) using ca. 3 ml ethereal diazomethane solution (Pressley 1982).
Excess diazomethane was removed by passing a gentle stream of nitrogen over the sample and
then the entire contents quantitatively transferred to a 11 x 300 mm glass column filled with 5
g florisil (60-80 mesh) topped with 1-2  cm anhydrous sodium sulfate. These prepared columns
were stored at 130°C for at least 24 hours prior to use.  Just before sample addition, each
column was rinsed with several  milliliters of 5 % methyW-butyl ether/hexane solution (elution
solvent).  Methylated TCG and surrogate were eluted with 40 ml elution solvent and collected
in a  100-ml volumetric flask.  Hexachlorobenzene internal standard solution (1.0 ml 880
/ig/liter) was  added and the contents diluted to volume.  Required dilutions were made with
internal standard solution at 8.8 /xg/liter.  Controls and matrix spike samples (controls to which
10 jig of TCG and surrogate had been added) were treated similarly.

       Nf3 electron capture gas chromatographic analysis - TCG and surrogate concentrations
were determined by gas chromatographic analysis using a Varian model 6000 gas
chromatograph equipped with a  Ni63 electron capture detector, autosampler and data system.
The column used was 180 cm x  2 mm i.d. glass column packed with 3% Carbowax-20M on
100-120 mesh Supelcoport™. This column was operated at 160°C.   Complete resolution of
internal standard, TCG, and surrogate was accomplished under these conditions.  Retention
times of hexachlorobenzene, surrogate,  and TCG were 7.6, 10.0, and 11.4 minutes.  Nitrogen
carrier gas was purified nitrogen at 25 ml/minute.  Detector and inlet temperatures were  set at
300 and 200°C, respectively.  Linear regression parameters from calibration curves were used
to quantitate toxicant concentrations.  Calibration standard concentrations ranged from 0.5-10
/xg/liters.  All samples and standards were injected in duplicate.  TCG levels were not
corrected for  surrogate recoveries.

TCG analysis - Water

       Two milliliters of each 5-minute respirometer water sample, and 5.0 ml of each 2-hour
sample, were placed in 100-ml square-shaped volumetric flasks containing   0.5 /xg surrogate,
ca. 50 ml double distilled water,  2 ml 18N sulfuric  acid,  10 ml  1:1 methyl-f-butyl
ether/hexane, and a magnetic stir. Extraction proceeded by mixing the  flask contents on  a
magnetic stirrer for >30 minutes.  Extraction solvent was then forced up into the neck of the
flask by addition of more water, and 1.0 ml of this  solvent transferred to a 10-ml volumetric

                                         57

-------
flask. Diazomethane solution (1-2 ml) was added and the solution let stand for at least 30
minutes. Excess diazomethane was removed, internal standard was added (1.0 ml at 88
/xg/liter) and the contents diluted to volume with hexane.  TCB and surrogate concentrations
were determined by gas chromatography as previously described.  Method blanks and TCG
(0.5 ng) fortified blanks were analyzed along with the water samples.
TCG method validation - Fish

       To ensure that quantitative recovery of TCG could be accomplished at concentrations
anticipated to be observed in the exposed fish, different amounts of TCG (5, 10, 20, 50, and
100 ^g) were added to each of a series of samples of small rainbow trout taken from the same
pool of trout used in these tests.  These fortified samples represented a TCG concentration
range in the  fish from ca. 1-24 /xg/g.  Surrogate (10 ^g) was added and then these samples
were taken through the same procedure used to analyze the small trout exposed to TCG in the
respirometer. TCG and surrogate recoveries were calculated.

Lipid analysis -

       Total lipids were determined (Folch 1957) in duplicate on each large fish tissue
homogenate. Lipid concentrations were also determined in the six small rainbow trout from
each of the two tests previously described. This procedure involved blending 5-g aliquots of
frozen tissue homogenate or diced small whole fish  (previously weighed) to a powder in a
small stainless steel blender, to which dry ice and anhydrous sodium sulfate were also  added.
Lipids were  extracted by blending this powder two times for 3 min each with 100 ml 2: 1
chloroform: methanol.  Two extractions were used.  The combined extracts were purified by
washing with aqueous potassium chloride solution (0.88 % w/v). Layers were allowed to
separate overnight and the aqueous layer discarded.  The organic layer containing the lipids
was vacuum filtered through a 0.2-pi Nylon-66 membrane and made to a 250 ml volume. A
25-ml aliquot was added to a tared beaker and solvent was removed using a gentle stream of
nitrogen and low heat.  The beaker and lipid residue was vacuum dried, weighed, and percent
lipid calculated.

                                      RESULTS

       TCG concentrations (wet weight basis) in largescale suckers ranged from 2.24-8.64
fjtg/g (Table  1).  When expressing this toxicant loading in terms of oxygen consumption, the
rate of uptake in these fishes varied from 5.94 - 23.99 ng TCG/
mg O2/hour  and averaged 12.43.  Concentrations of TCG in  large rainbow trout measured
6.76 - 13.01 fjig/g with uptake rates calculated at from  8.49 - 12.90 (mean 10. 17).  Small
rainbow trout had TCG concentrations ranging from 13.74-21.97 /*g/g with corresponding
uptake rates  of 9.51 - 20.15 fig TCG/mg O2/hour (mean 13.31).

       Duplicate TCG values measured in the homogenate samples agreed within 10% for all
except three samples where differences were noted as high as 16%. Surrogate recoveries in
these tissue homogenates ranged from 84.5 - 119%. TCG recoveries from fortified control
                                        58

-------
Table 1.  Test summary
DOSE
w/i
Largescale SLK
215
208
223
208
215
200
215
208
215
215
mean
Large rainbow
261
261
261
261
258
261
261
258
258
255
mean
Small rainbow
256
H
II
H
M
II
II
M
259
n
«
"
n
H
II
H
mean
5min-TCG
M/l
•kers
244
230
220
211
210
189
214
194
196
203

trout
224
227
230
240
239
223
227
229
234
229

trout
253
M
M
II
M
H
II
II
256
n
"
H
«
II
II
II

H20 T
°C

9.15
9.19
9.20
9.23
9.31
17.97
18.09
17.94
18.08
17.97


7.10
7.08
7.15
7.06
7.06
7.16
7.09
7.13
7.08
7.17


5.00
H
H
It
II
H
II
H
4.90
n
H
H
II
II
II
H

FISH
nb

SFU-1
SFU-2
SFU-3
SFU-4
SFU-9
SFU-5
SFU-6
SFU-7
SFU-8
SFU-10


223
225
226
231R
232
224
227R
228
229
230


a-1
a-3
a-9
a-13
a-18
a-20
a-24
a-27
b-8
b-10
b-12
b-15
b-20
b-21
b-22
b-24

FISH UT
g

668
931
745
780
685
553
868
582
850
403
707

464
493
487
360
382
392
446
367
344
360
410

4.74
4.66
5.40
5.68
4.90
5.22
4.80
5.39
4.83
5.32
5.66
5.78
5.38
5.69
5.82
5.62
5.31
LIPID
*

7.96
9.23
8.19
4.47
9.78
10.83
11.22
3.21
7.84
8.06
8.08

6.25
7.67
7.02
9.13
7.36
7.42
9.64
9.52
8.21
6.14
7.84

(1)
H
II
H
If
n
n
H
u
n
M
II
II
M
M
H

SPEED
Bt/s

.13
.09
.07
.12
.11
.01
.07
.12
.08
0.99


1.40
.30
.32
.29
.26
.88
.71
.95
.90
.90


2.41
2.44
2.29
2.23
2.38
2.35
2.38
2.29
3.48
3.44
3.35
3.48
3.31
3.40
3.35
3.31

H02
rag/h

56.79
53.99
50.86
34.85
64.55
71.88
123.69
87.04
118.68
47.55


75.2
102.0
93.3
84.1
104.0
106.0
104.0
133.0
108.0
92.6


1.27
1.27
1.46
1.54
1.33
1.40
1.30
1.46
1.88
1.99
2.13
2.13
2.02
2.09
2.17
2.09

2 M02
mg

113.58
107.98
101.72
69.70
129.10
143.76
247.38
174.08
237.36
95.10


150.4
204.0
186.6
168.2
208.0
212.0
208.0
266.0
216.0
185.2


2.54
2.54
2.92
3.08
2.66
2.80
2.60
2.92
3.76
3.98
4.26
4.26
4.04
4.18
4.34
4.18

TCG
M/g

3.52
4.26
6.55
2.48
2.24
4.10
6.70
8.64
5.68
3.57


8.36
7.03
6.76
10.08
10.86
10.88
11.46
13.01
12.06
10.11


17.09
21.97
15.78
15.77
18.55
13.74
17.27
15.98
19.98
14.23
17.70
16.59
16.26
15.49
15.91
14.42

2 TCG
M

2351
3966
4880
1934
1534
2267
5816
5028.
4828
1439


3879
3466
3292
3629
4149
4265
5111
4775
4149
3640


81.01
102.38
85.21
89.57
90.90
71.72
82.90
86.13
96.50
75.70
100.18
95.89
87.48
88.14
92.60
81.04

RATE
A/rag/h

10.35
18.36
23.99
13.88
5.94
7.89
11.75
14.44
10.17
7.56
12.43

12.90
8.49
8.82
10.79
9.97
10.06
12.29
8.97
9.60
9.83
10.17

15.95
20.15
14.59
14.54
17.09
12.81
15.94
14.75
12.83
9.51
11.76
,1.25
10.83
10.54
10.67
9.69
13.31
   (1)  TCG analysis required used of the entire fish so lipid levels could not be measured on these fish.
       Lipid concentrations were measured on 6 additional fish from each test group (range 3.44 - 7.57X,  mean = 5.75).
                                                       59

-------
fish homogenates varied from 84.5 - 107%.  No TCG was detected in any of the control fish
analyzed.  Surrogate recoveries from the small rainbow trout tested ranged from 86.9 - 108%.
Recoveries of TCG in untreated small rainbow trout varied from 86.4 -  108%.

       TCG concentrations in the 5-minute respirometer water samples collected during each
test ranged from 189 - 256 /xg/liter indicating proper toxicant addition at the start of test.
Surrogate recoveries from these samples ranged from 89.7 - 1 15%. TCG recoveries from
fortified control water samples and method blanks varied from  95.6-109%.  No TCG was
detected in the control water sample or in the method blanks analyzed.

       Lipid concentrations in largescale suckers varied from 3.21 - 11.22% total body
weight, and averaged 8.08%.  Lipid concentrations in large rainbow trout ranged from 6. 19 -
9.64% (mean 7.84%).  Variance in duplicate analysis of the tissue homogenate samples was
never greater than 10%. Lipid concentrations in small rainbow trout lipid values ranged from
3.44 - 7.57%,  averaging 5.75%.

       Recoveries of TCG in the control fish fortified at various levels ranged from 86.4  -
102% with a mean calculated at 94.12 (Table 2).  Linear regression analysis comparing
analytical values to fortification concentrations indicated quantitative recovery over the
concentration range tested (Figure 1):
                                   - 0.297    (r2 = 0.9974)

Surrogate recoveries in these samples ranged from 88.3 - 105%.

                                    DISCUSSION

       Mean TCG uptake rates of the three groups of fishes tested were very similar. The
small rainbow trout absorbed the greatest amount of toxicant at 13.31 /xg TCG/mg O2/hour,
followed by the largescale suckers at 12.43, and then the large rainbow trout at 10. 17.  When
the uptake rates of all the fishes tested are evaluated as  a single group, the mean uptake
(n=36) is 12.93 /xg TCG/mg O2/hour with a 95% confidence interval of ± 2.53.  These
values imply that all the fishes essentially absorbed TCG at about the same rate and indicates a
direct correlation between waterborne TCG uptake (at a given water concentration) and oxygen
consumption by the fish. This relationship is shown graphically in Figure 2.  More data are
required to develop a predictive absorption model. Tests should involve medium size rainbow
trout, medium and small largescale suckers, and then other species of fishes.

       Lipid concentrations in these fishes were also very similar, especially in the larger
fishes.  However, even though lipid concentrations were similar, the mean TCG concentration
in the largescale suckers was slightly less than half that of the larger rainbow trout, but at the
same time their mean oxygen consumption rate was approximately 70 % of that of the larger
rainbow trout (Table 1).  This indicates the important role of oxygen consumption rates in
rates of toxicant bioconcentration. When uptake rates  of these two groups of larger fishes
were tested for significance using the Mann-Whitney (Wilcoxon) nonparametric rank sum test,

                                         60

-------
  -a
  "CT
     25 r
     20
  ±r  15
   (0
   o  10
                                          y = 1.02 x  - 0.297


                                          r = 0.9987
                             1O          15



                           Fortification
                                                   2O
                                                              25
       Figure 1.  TCG recovery from fortified rainbow trout,

                        method validation
   10000
£t

-------
 Table 2.  TCG fish analysis - method validation.
      FISH
                WEIGHT (g)
                          (ig TCG ADDED
                                       H3/3 (THEORETICAL)
M9/9 (ANALYTICAL)
                                                                      X RECOVERY
C-14
C-11
C-10
C-18
C- 5
C-15
6.22
5.57
5.90
5.09
6.48
4.22
0
5
10
20
50
100
0
0.898
1.69
3.93
7.72
23.7
0
0.892
1.51
3.68
6.67
24.2
.
99.3
89.3
93.6
86.4
102
no differences could by measured.  This may not be a valid test with so few numbers;
however, one-way analysis of Variance (ANOVA) applied to all three groups of uptake rate
data suggested that the differences in the means of the groups was no more significant than the
variability within each group.

                              ACKNOWLEDGEMENT

      This research was funded in part by the U.S. Environmental Protection Agency,
Environmental Research Laboratory, Athens, Georgia, through Cooperative Agreement
CR 995189.

                                   REFERENCES
Dence, C. W.  1971. Reactions of lignins, halogenation and nitration. In: Lignins,
      occurrence, formation, structure and reactions. K.V. Sarkanen and C.H. Ludwig
      [ed.].  John Wiley and Sons, Inc., New York, N.Y.

Folch, J., M. Lees, and G.H. Stanley.  1957.  A simple method for the isolation and
      purification of total lipides from animal tissues.  Journal of Biology and  Chemistry
      226: 497-509.

Gehrke, P.C., L.E. Fidler, D.C. Mense, and D.J. Randall.  1990.  A respirometer with
      controlled water quality and computerized data acquisition for experiments with
      swimming  fish. Fish Physiology and Biochemistry 8: 61-67.

Landner, L., K. Lindstrom, M. Karlsson, J.  Nordin, and L. Sorensen. 1977.
      Bioaccumulation in fish of chlorinated phenols from kraft pulp mill bleachery
      effluents.  Bulletin of Environmental Contamination and Toxicology 18:  663-673.

Leach, J.M., and  A.N. Thakore.  1973. Identification of constituents of kraft pulping
      effluent that are toxic to juvenile coho salmon (Oncorhynchus kisutch).  Journal of the
      Fisheries Research Board of Canada 30: 479-484.
                                        62

-------
Leach, J.M., and A. N. Thakore.  1975.  Isolation and identification of constituents toxic to
       juvenile rainbow trout (Salmo gairdneri) in caustic extraction effluents from kraft pulp
       mill bleach plants. Journal of the Fisheries Research Board of Canada 32: 1249-
       1257.

Niimi, A.J., H.B. Lee, and G.P. Kisson.  1990.  Kinetics of chloroguaiacols and other
       chlorinated phenolic derivatives in rainbow trout (Salmo gairdneri).  Environmental
       Toxicology and Chemistry 9: 649-653.

Paasivirta, J., P. Klein, M. Knuutila, J. Knuutinen, M. Lahtipera, R. Paukku, A.
       Veijanen, L. Welling, M. Vuorinen, and P. Vuorinen.   1987.  Chlorinated
       anisoles and veratroles in fish.  Model compounds. Instrumental and sensory
       determinations.  Chemosphere 16:  1231-1241.

Pressley, T.A., and I.E. Longbottom.  1982.  The determination of chlorinated
       herbicides in industrial and municipal wastewater - Method 615.  Environmental
       Monitoring and Support Laboratory, U.S. Environmental Protection Agency,
       Cincinnati, Ohio.

Renberg, L., O. Svanberg, B. Bengtsson, and G. Sundstrom.  1980. Chlorinated
       guaiacols and catechols bioaccumulation potential in bleaks (Alburnus alburnus,
       Pisces)  and reproductive and toxic effects on the harpacticoid Nitocra spinipes
       (Crustacea).  Chemosphere 9: 143-150.
                                         63

-------
          AN INVESTIGATION OF THE LOWER DON RIVER WATER QUALITY
                            USING SIMULATION MODELS
                                         by
       M. G. Yereschukova1, E. Z. Hosseinipour2, R. B. Ambrose3, V. V. Tsirkunov1,
                          R. C. Russo3 and A. M. Nikanorov1
                                    ABSTRACT

      This paper describes a hydrodynamic and water quality modeling effort on the Don
River in  Russia. The waterway is one of the largest rivers of Russia and has been adversely
impacted by development; its lower part in particular is severely impacted by anthropogenic
processes.  Don River basin is an intensively exploited region with respect to agricultural and
industrial activities. Waters of the Lower Don system are diverted for municipal and industrial
uses, as well as for irrigation of agricultural fields.  Preliminary results from water samples
collected in the summer of 1990  indicated that conventional pollutants and synthetic organic
materials are the major contaminants of concern.  The developed linked model of the system
consists of a river hydraulics and sediment transport code (RIVMOD),  its modified version for
handling branched systems (RIVNET), and the water quality modeling package (WASP4).
Simulation results indicate that detailed morphometric characteristics are needed at the Don
mouth for better model performance.  Hydrodynamic and water quality data should also be
collected on the main branches of the Don at its mouth  on a concurrent basis in order to be
useful for modeling purposes. To determine the best management alternative for water
quality improvement, watershed models should  be used to delineate the nonpoint source
loading patterns.   For these purposes, information on water uptake from the river and
wastewater loading to the river are also required.  Although evaluation of model performance
is difficult due to inadequacy of information and uncertainty about representativeness of
available data, nevertheless results are useful for designing data collection networks for
long-term water management and further simulation studies.
1Hydrochemical Institute, Rostov-On-Don, Russia.
2South Florida Water Management District, Planning Department, West Palm Beach, FL
 (USA).
3U.S. Environmental Protection Agency, Environmental Research Laboratory, Athens, GA
 (USA).
                                        64

-------
                                   INTRODUCTION

       Among the large rivers of Russia, the Don-and in particular its lower part-is severely
affected by anthropogenic impacts. Runoff and irrigation return waters from fields,
discharges from industrial plants along the river, and discharges from municipalities adversely
impact the quality of Don waters. Although there are heavy industrial complexes in the river
basin, preliminary data from water samples showed that conventional pollutants and synthetic
organic materials are the major  contaminants of concern. For water quality modeling,
information about hydrological regime, transport and transformation of contaminants carried
to the river, and river ecosystem characteristics are required.  Such data and associated
simulation results from modeling can then provide necessary inputs for water management
alternatives that are environmentally and ecologically sound and economically feasible. To
this end,  hydrodynamic and water quality modeling of the Lower Don River system can shed
some light on the processes active in the waterway and  provide insights that could be helpful
in selection of water quality management alternatives.

       In this paper only the stretch of the Don River from Razdorskaya to Taganrog Gulf is
considered in the simulations. Analysis of data collected by the Hydrochemical Institute in
Russia, and the Environmental Protection Agency's Athens Environmental Research
Laboratory (AERL) in the U.S., showed that  the main water quality problems of the Lower
Don system are associated with processes of eutrophication, contamination by organic
matter, and pesticides. Besides, high sediment load and turbidity may essentially influence
the aquatic ecosystem and hence self-purification processes.
                                 PHYSICAL SETTING

       The Don region covers an area of about 323,000 km2.  The Don watershed consists of
three different physiographic zones: mixed forest, partially wooded steppe, and steppe. This
region  is part of the Don River watershed, supplying much of its annual flow.  The Don River
is situated between 44 and 54°N latitude and 37-45°E longitude. The maximum north-south
spread is 650 km, and the east-west spread is about 160 km.  The Don is a relatively large
river, approximately 1870 km long and is fed by a watershed with an area of about 422,000
km2. The river's  headwaters are on the east side of the Middle Russian Hills. The Don flows
into the Taganrog Gulf on the Azov Sea. In the upper parts, the river flows through a  narrow
valley and its channel has high right banks.  In these parts the river also intercepts many
ravines and has a number of shoals, which are formed by the sediment transport from
tributaries and inflows from ravines. The river  channel is very winding and often meanders in
the flat sections.

       In the middle section the river valley is significantly wider, though the river channel
runs close to the right side of the valley with elevated banks. At the end of the Don's middle
section the natural flow is altered by a hydropower and water supply dam behind which is the
Tsimlyansk Reservoir. Downstream from the dam the river valley is quite wide with lateral
stretches up to 30 km. Water depth in the river channel at the thalweg can be as much as
20 m.  The river estuary is situated downstream from the city of Rostov-On-Don.  The  estuary
covers an area of approximately 340 km2. In the lower reach of the river  more tributaries
discharge into the Don.
                                        65

-------
       Anthropogenic activity plays a significant role in transformation of natural features of
the Don River. The river basin is one of the most exploited regions of the country with
respect to industry and agriculture.  Ecological problems such as soil plugging, forest cutting,
and soil erosion are widespread.  In addition to the  Tsimlyansk Reservoir on the Don,  a
number of reservoirs also are in place on the Don's tributaries, as well as hydraulic
engineering structures such as dams and irrigation channels.  Almost 7500 km2 of agricultural
fields are irrigated in the Don River basin.  The Don River and its tributaries are sources of
water supply for many large industrial complexes. The stretch of the Don River from the
Tsimlyansk Reservoir to the Taganrog Gulf is referred to as the  Lower Don River. The area
of the watershed in the lower part is over 160,000 km2; the length is about 313 km.  The river
in the lower part meanders through the valley with a spread of usually about 11-12 km,
though it can  reach  up to 22-25 km.  The river width varies from 400 to 600 m in the lower
part. Average water depth during low-water regime is in the range of 4-6 m in the main
channel, and  decreases to 0.7 m in  shoal areas.
                 HYDROLOGY OF THE LOWER DON RIVER SYSTEM

      The Lower Don region is situated in a semiarid climate zone with characteristics of
which are irregular precipitation and highly variable surface and ground waterflow.  The main
source of water is precipitation. Average total runoff is estimated to be about 30 mm per
year of which 20 mm is produced by surface runoff and the rest from  ground water runoff as
interflow or baseflow. Direct surface runoff to the river from precipitation produces 75% of
the flow of which snow accounts for about 67%. The rainfall component of surface runoff is
no more than 8% of whole river flow. Hence, the majority of the river discharge takes place
in spring  during snowmelt. Overall, summer rains and winter snow characterize the
hydrology of the Lower Don River system and contribute  most of the annual flow. The
hydrochemical composition of the Lower Don River is influenced by many factors. The
primary chemical characteristic and biological features are defined by the waters behind the
Tsimlyansk Reservoir. As water flows downstream, its chemical composition transforms
significantly. This transformation is due to loading of the  dissolved matter which  is
anthropogenic and of natural origin. Yereschukova et al. (1992) give  more details on the
river system.

                          WATER QUALITY CONSTITUENTS

SEDIMENT WASHOFF AND ITS CHARACTERISTICS

      Most sediment washoff to the river occurs during spring high flows.  Suspended
sediment distribution across the channel cross-section is  both temporally and spatially very
heterogeneous.  Turbidity increases in the middle sections of the river and near the bed.
Surface  soil layers in the watershed of the Lower Don River have a highly impermeable crust.
This feature causes  rapid accumulation of water from rainfall, since very little water can
infiltrate the soil crust. Runoff carries with it all the accumulated fine materials from the soil
surface.  The concentrated flow can harness enough energy to induce erosive power to the
point of gully erosion and heavy sediment load to the river system.
                                        66

-------
       During spring more than 70% of annual solid loads are transported to coastal waters.
The annual load of suspended matter is close to 1.6 million tonnes in the Lower Don. About
40% of these materials are accumulated in the river channel; only 1 million tonnes are
transported to the coastal-sea areas (Yereschukova et al. 1992).  The composition of
suspended materials in the lower part of the river may provide some hint concerning
conventional water quality parameters. The  transported sediments consist of mineral matter,
detritus and live bacteria, phytoplankton, and zooplankton.  Analysis of transported materials
showed that during spring 1984 solid load comprised 24.3% and live organisms and  minerals
75.5% of the  total load.  The water column saturation with mineral load has negative effect
on the biota of the Lower Don. Phytoplankton is a source of significant biomass;  composition
of different species is sufficiently stable. During spring much of the phytoplankton consists of
diatoms (61-93% of total biomass). Also represented are the greens and goldens which are
about 2-29%  and 0.08-39%,  respectively; bluegreens do not play a very significant role (1-
17% of total biomass).  Quantitative analysis indicates that phytoplankton distribution is very
uneven.  Total biomass changes from 0.5  to 6.8 mg/L, with an average value of 3.0 mg/L.
DISSOLVED OXYGEN REGIME AND OTHER CONTAMINANTS

       The dissolved oxygen regime of the river is fairly stable. Oxygen content has never
decreased below critical level (i.e., 5 mg/L).  At the same time, in Tsimlyansk Reservoir,
during intensive growth of blue-green algae anoxic conditions may  prevail and the oxygen
concentration in the bottom layers can  go down to 1 mg/L.  Concentrations of heavy metals,
phenols, and synthetic surfactants are generally stable within the study area and slightly
exceed maximum allowable concentration value.  Concentrations of chloro-organic pesticides
such as DDT, lindane, and their metabolites are also measured for routine monitoring
purposes.  These pesticides were detected in more than 50% of the samples for all locations.
Maximum concentrations of DDT may reach dozens of micrograms per liter, while lindane
concentrations of several hundred parts per billion are not uncommon.  Determinations of
concentrations for trace organic pollutants showed that detectable concentrations were
predominantly in zones of direct influence from major point sources of pollution.  Most of
these pollutants are biodegradable, so they may have only  local toxic effects.
NUTRIENTS

       Part of the chain in the nutrient recycling process is dissolved organic matter. The
amount of dissolved organic matter is measured by the amount of oxygen needed for
oxidation of the organic load or the biochemical oxygen demand. The concentration of
dissolved organic matter is highly variable.  In 1984 it ranged from 4.32 up to 15.52 mg O2/L.
The organic load distribution along the river is uneven.  There is a small increase in the
amount of dissolved organic matter from Tsimlyansk Dam to the mouth of the Don (6.08 to
7.2-11.52 mg O2/L). This trend is disturbed only at the mouth of tributaries Sal and Manuch.

       There  is a clearer trend in distribution of organic matter in the Don across the river
section. Higher concentrations have been observed at the right bank of the river.  The
reason is that more sources of anthropogenic pollution are situated on the right bank of the
river.  The nutrient load to the Lower Don River system changes unevenly during the year
                                         67

-------
and from 1  year to the next.  Nutrients come to lower parts of the Don River from upstream
tributaries and lateral point and nonpoint sources.  The average load through the
Rasdorskaya cross-section is presented in Table 1.

      Sources of the phosphorus supply to the Lower Don are varied.  Part of it is due to
decay of dead organisms; another part is supplied through bottom erosion processes.  A
significant portion  of phosphorus load can be attributed to return water from agricultural
fields, stock farms, and municipal wastewater treatment plants. During spring the mean
concentration is about 0.009 mg/L (with a range of 0.006-0.012).  In spring, concentrations of
phosphorus in the Lower Don River are at background  levels.  In summer, concentrations rise
to very high levels, sometimes up to 0.3 mg/L, and the  waterbody shows signs of low level
phosphorus contamination. In autumn, phosphorus concentrations decline to 0.09 mg/L.
TABLE 1. NUTRIENT LOADS TO THE LOWER DON RIVER SYSTEM (IN METRIC TONNES PER
YEAR)
Month
NO,
NO,
NH4
Si
January
February
March
April
May
June
July
August
September
October
November
December
830
1120
1770
2550
3630
3060
1010
1090
1860
1450
3080
1770
74
88
223
78
147
162
42
60
217
32
70
26
400
590
85
1670
1480
60
250
360
290
320
460
280
8
29
60
105
76
53
41
50
50
46
59
25
4
5
8
10
10
6
4
7
5
7
8
5
             WATER QUALITY PROBLEMS IN THE LOWER DON SYSTEM

       Based on data from observations, water quality problems of the Lower Don can be
characterized only in general terms. For example, the high hydrocarbon level can be
explained by intensive use of the river as a waterway for navigation and transportation.
Petroleum products, such as gasoline  and diesel fuel, may leak from small boats and ships
used for transport and recreational purposes and from underground storage tanks in the
vicinity of the river;  these are considered major sources of hydrocarbon contamination.  In
addition, more hydrocarbon  loadings may be added to the river from urban runoff.  Partially
treated wastewater from domestic sources through sewage treatment plants or by direct
discharge may be contributing to high  nitrite concentrations.  Detectible pesticide
concentrations can  be interpreted as a result of intensive  application on crops and other
plants in the surrounding watersheds.
                                        68

-------
                            THE MODELING FRAMEWORK

       The above brief description of characteristics of the Lower Don River system shows
that, in order to get a feel for the enumerated problems, models that can describe the
hydrodynamic, sediment transport, and water quality processes in branched river systems are
needed.  The most appropriate models for this kind of study are those with dynamic riverine
water quality simulation. However, most available models did not have the capability for
sediment transport or handling branched systems.  Because of familiarity and previous
experience we used the WASP4 water quality modeling package (Ambrose et al. 1988) in
association with a river hydraulic and sediment transport model, RIVMOD (Hosseinipour and
Martin 1991). The water quality model of the system is based on mass balance in a
compartment of a waterbody.  Hydrodynamic parameters for the  water quality model were
supplied by a modified version of RIVMOD for branched systems called RIVNET. The model
is based on the one-dimensional equations of continuity and momentum for a branching or
channel-junction computations network. The resulting unsteady hydrodynamic information
from RIVNET are averaged over large time intervals and stored for later use by the water
quality component of the simulation package. Simulations were conducted for spring high
flows, considering branches and creeks flowing into the system.
                           WATER QUALITY SIMULATIONS

       Hydrochemical processes play an important role in mixing and transport of chemical
and biological compounds in natural waterbodies.  To simulate hydrochemical and
hydrobiological processes in rivers, discharges, velocities, and depths are needed.
Therefore, water quality modeling of the Lower Don River system was started with
hydrodynamic simulation.  At the upstream of this part of the river, daily discharge and stage
are available, but these data belong to the period before  1987.  Besides, short-term
variations in flow may not be significant in long-term simulations. Thus, average values of
discharge for the 10-day period were chosen as a boundary value.  Downstream boundary
values for discharge or depth can be specified only approximately because very little data on
the branches at the mouth of the Don River exist.  Lack of precise information induces
increasing uncertainty about modeling results. To estimate the influence of errors with regard
to boundary conditions on results of simulation, an uncertainty analysis was conducted
(Yereschukova et al.  1992).  To compare results of different experiments, the measure of
deviations between base runs and solutions with modified input data sets was introduced as
follows:
                                                       ]                          d)
                                /V /=1        /V /=i

where
       A            = absolute measure of error influence
       Xb(t,)         = the value of characteristic at time t,, which is obtained from the
                          simulation with the basic .input data
       X(t,)          = the value of characteristic at time t,, which is obtained from the
                          simulation with modified input data
       N            = the number of time steps during the simulation.
                                         69

-------
The relative measure is a more representative estimator of uncertainty. The relative measure
can be calculated by the following formula (Jorgensen 1 986):
                                          A
                                      1   N
                                     ~ £
                                      N M                                         (2)
                                 6=   - 100%
where
       6    =       the relative measure of changes in %
       Ap/p =       the relative change in the parameter value (here in the boundary value).

       The above formulas were applied to estimate the change in water distribution between
branches, which are caused by a change in boundary conditions.  Such information is very
important in identification of water quality parameters for modeling. To provide accurate
simulation of the hydrodynamic regime at the mouth of the Don River, downstream boundary
conditions have, to be specified as accurately as possible, and no less accurately than
upstream boundary conditions. This accuracy is even more important for simulation of low
flow events. The last cross-section where discharge is measured  is near Razdorskaya.  The
discharge does not change significantly along the river channel; the high discharge period
coincided with the beginning of the study, and  after that their deviations from average flow do
not exceed 15% of mean value.  Hence, it is possible to use constant discharge for water
quality simulation. An example of results of discharge simulations from  May to October 1983
are presented in Fig. 1 .

       As stressed above, one of the main water quality problems in the Lower Don River
system is eutrophication. Simulations could help to detect crucial parameters affecting the
river ecosystem. Once those parameters are determined,  the trophic level of this waterbody
can be estimated more precisely. Therefore, the model used for this purpose  has to contain
the stated variables at least for the description of algal biomass and organic and inorganic
nutrients.  During preliminary analysis of available water quality data, we found that the data
are not adequate for model calibration.  However, calibration of the model, even with
incomplete data, could be useful for exploration of the main stated variables and other
parameters of interest.  As a whole, simulation results can hardly be used for testing different
management strategies for water quality improvement in the region.  However, simulations
can help determine areas that need more research and plan for data collection and further
investigations.

       Eutrophication simulation was conducted using the EUTRO4 module of the WASP4
modeling package. This package allows several choices for eutrophication modeling.
Considering the fact that adequate data were not available to describe dynamics of different
phytoplankton groups, the following stated variables were included in the model: biomass of
total phytoplankton, organic nitrogen and phosphorus, and inorganic phosphorus and nitrogen
constituents (including ammonium and nitrate).  Although diatoms  constitute a large part of
spring  phytoplankton and silicon is the nutrient that affects their dynamics, this nutrient cycle
is not taken into consideration because it is not taken up by the other groups of algae.
                                         70

-------
                               Rostov-on-Don
           700




           650




           600




           550




           500,
                 May
              June      July     August   September  October
   O
   UJ
   CO
   UJ
   O
   O
   (0
   5
   £
   111

   i
280



260


240


220



200



180



160
                                      Azov
I   I	1_
           420


           410

           400


           390

           380

           370


           360


           350


           340
      May      June      July    August  Septeirber  October
                                    Dug 1 no
    j	I	|	I
                   __i	*	*	I	i
                 May
               June
                    July
August  September  October
Fig. 1.  Water discharges in the Don mouth.
                                    71

-------
Implicitly, the water quality model assumes that silicon is never depleted enough to limit
diatom growth.  Water temperature and light intensity significantly affect phytoplankton
dynamics. These parameters are included in the model as forcing functions.
                               SENSITIVITY ANALYSIS

       Sensitivity analysis is an important step in environmental transport and fate modeling,
especially in situations where data are incomplete and/or unreliable.  Such analysis allows
researchers to look into the most crucial parameters of the model and get a feel for their
impact on the overall analysis. To quantify the influence of parameter changes on dynamics
of the stated variables, equation  2 was applied. Values were varied +10% of their base
values.  The analysis was conducted according to a 1-factor plan.  Maximum values of
relative changes based on equation 2 are presented in Table 2.
TABLE 2. MODEL SENSITIVITY DUE TO PARAMETER VARIATIONS. PARAMETERS AFFECTING
PHYTOPLANKTON  DYNAMICS ARE MAXIMUM GROWTH RATE (G), RATES OF DECOMPOSITION
OF ORGANIC PHOSPHORUS AND NITROGEN (Kp, KN), MORTALITY RATE (M), LIGHT SATURATION
COEFFICIENT (L), AND CARBON-TO-CHLOROPHYLL RATIO (r)

Variable                     G        Kn       KN       M          L          r
Phytoplankton
Ammonia
Nitrate
Organic nitrogen
Inorganic phosphorus
Organic phosphorus
0.0850
0.1641
0.0237
0.0438
0.0248
0.0041
0.0014
0.0018
0.0006
0.0005
0.0616
0.3633
0.0000
0.3643
0.0601
0.0000
0.0000
0.0000
0.0220
0.3800
0.0592
0.0029
0.0010
0.0219
0.9444
0.1810
0.0259
0.0484
0.0272
0.0045
0.9991
0.1178
0.0283
0.0340
0.0185
0.0729
       It should be noted that the model is structurally stable in the domain of parameter
values in Table 2, because the values of criterion do not exceed 1.0.  Since phytoplankton
dynamics are most sensitive to variations in carbon-to-chlorophyll ratio, it would be useful to
determine more precisely the species composition of algae and their elemental composition.
Among factors affecting phytoplankton dynamics, light intensity plays the most important role
and is probably the most limiting factor on this ecosystem. Nitrogen influence on
phytoplankton dynamics is not very significant, in part because its concentrations in Lower
Don waters are significantly higher than its half-saturation value. Changes in phosphorus
dynamics do not affect significantly the algae biomass, because its concentrations are not
significantly below half-saturation values.

       Effect of waterflow on the modeled ecosystem was investigated.  Flow monitoring data
and hydrodynamic simulations indicate that the discharge does not fall below 400 m3/sec in
the Lower Don system during the growing season. At times the discharge may reach as high
as 1800 nrvVsec in the beginning of spring (e.g.,  1979), but after a short time it decreases to
about 500-600 m3/sec. A discharge of 600 m3/sec was chosen as the baseflow. The model
                                         72

-------
was run with the same input data set, except discharges were changed to 400 and 1000
m3/sec (see Fig. 2). It appeared that the maximum concentration of phytoplankton biomass
does not change significantly with flow.  Values were close to those at the upstream
boundary.  Even increasing travel time when discharge  is low did not cause intensive
blooming of phytoplankton.  We attempted to estimate the influence of flow redistribution
between branches on  phytoplankton dynamics at the mouth  of the Don River. As described
above, significant changes in distribution of flow can be  caused by errors in  specifying
hydrodynamic boundary conditions.  During water quality simulations, however, only slight
deviations in biomass  (2% of maximum value) were discovered.  It should be noted that only
part of the Don River mouth was modeled, and travel  time along the branches did not
increase sufficiently to allow phytoplankton to bloom.

       Observations indicate that phytoplankton distribution is very uneven along and across
the river. Near the banks, concentrations of phytoplankton biomass may be very high.
Simulation of this phenomenon with a 1-D model is difficult and because hydrological and
hydrodynamic data are sparse, 2-D modeling is not feasible.  It is possible, nevertheless, to
estimate probable effect of phytoplankton spots on average phytoplankton concentrations on
this reach of the river under certain assumptions. It was assumed that the five upstream
segments of this part of the Don River are connected  with shallow pools, where
concentrations of phytoplankton reach their maximum values of nearly 6-7 mg/L.
Phytoplankton from these pools may be transported to the main channel by  dispersion.
Simulations  including this extra load of phytoplankton did not result in significantly higher
phytoplankton biomass concentrations in the main channel.  Maximum concentrations
increased only about 7%. Hence, a major part of phytoplankton biomass comes from
upstream, and is not significantly augmented by phytoplankton that may disperse into the
main river channel from shallow waters near the banks.
                                   MODEL TESTS

       Hydrologic data on water stage applicable in modeling are available only on the
segment of the river from Razdorskaya to Rostov-on-Don. Therefore, simulations were
conducted for this part of the river.  Hydrodynamic model parameters were identified for July
1-10,  1979, and the model was run to simulate the system for this period.  In other tests, the
model was run with an input data set corresponding to February  5-15 and September 1-10.
Modeling results compared well with observed data; the difference between simulated and
measured values of stages did not exceed 5%.  The period of simulation, corresponding to
March 1-10,  1979,  seemed to be a good start.  During this time of year snow starts to melt
on the watershed of the Lower Don River system, and lateral inflows are a  significant part of
the river total flow.  Simulations were conducted with zero lateral  inflows to the segments,
because the real values of lateral inflows were unknown. As a result, simulated values of
stages were less than observed values by about 30% in the  middle segments of the river.

       Sediment transport simulation was conducted with hypothetical data because  of lack
of real data.  The model was run  for data corresponding with particle sizes  from 0.07 to 0.20
mm.  Results of simulation  showed that the river capacity for sediment transport is very
uneven from segment to segment.  The capacity is high for the marginal segments, and lower
in the middle portion of the river.  Results conform with the fact that shoaling is  proceeding
                                        73

-------
 I
                                      o
                                      o
                                      o
                                      •cH
                                      II
                                      a
                                      o
                                      o
                                                0)

                                                o

                                                o
                                                o
                                                t-l
                                                0)
                                                ^1

                                                0)
                                                -M
                                                PL,
                                                0)
                                                en
en
J3
tn
                                      o
                                      o
                                                 0)

                                                 Pi
(H             f
          CM

          en
                        74

-------
very intensively in the Lower Don River system. Sediment transport capacity depends on the
size of the particles.  Sand, with diameter of particles about 0.07 mm, is transported with the
discharge from 0.002 up to 0.036 rrvVsec.  For particles with a diameter of 0.20 mm, the
discharges are significantly less and vary from 0.004 up to 0.005 m3/sec. Sediment
discharges in shallower branches are higher than the deeper ones by about 2.5 times.
                                   CONCLUSIONS

       Modeling results and experimentation with simulations on the Lower Don River system
led to come conclusions that are helpful in planning future investigations.  First of all, more
detailed morphometric information  on the'mouth of the Don River is needed.  Additional
information is required concerning  channel segment lengths, depth, widths, and bottom and
surface slopes of branches and creeks in the delta and particularly near the junction of the
Don with inflowing and outgoing branches at the mouth.  Discharges and/or water stages
need to be measured at the junctions as well. Also, stage or discharge near the Taganrog
Gulf have to be measured with the same frequency as those at the upstream of the Lower
Don River system for boundary condition specification.

       Simulation of eutrophication processes showed that the main factor controlling
phytoplankton dynamics is the river discharge.  The river flow is sufficient to inhibit  intensive
phytoplankton blooms in the portion of river from Razdorskaya to the Azov Sea.  According to
the data, most phytoplankton biomass passing through this reach comes from upstream.
These  primary water quality simulations allow us to draw only general conclusions about
some features of the ecosystem.  To estimate the  eutrophication level at the Don River
mouth, the stretch of river from Dugino to the Taganrog Gulf has to be studied in more detail.
For this purpose, information about the segment should be collected as well as about the
whole Lower Don River system.  For a  comprehensive study, the data collection plan should
include gathering information about phytoplankton  dynamics  and nutrient cycle with at least 2
week frequency and at 10 km intervals  along the river. Quantitative  information about
phytoplankton species composition can also be very useful.  Additional study of macrophyte
distribution and their role in the eutrophication process would be very helpful in simulating
phytoplankton and nutrient dynamics more accurately.
                                   REFERENCES

Ambrose, R. B., T. A. Wool, J. P. Connoly and R. W. Schanz.  1988.  WASP4, a
       hydrodynamic and water quality model-model  theory, user's manual, and
       programmer's guide.  U.S. Environmental Protection Agency (U.S. EPA), US
       EPA/600/3-87/039, Athens, GA.

Hosseinipour, E. Z. and J. L. Martin.  1991.  RIVMOD-a one-dimensional hydrodynamic and
       sediment transport model. Model theory and user's manual.  South Florida Water
       Management District, West Palm Beach, FL.
                                        75

-------
Jorgensen, S. E. 1986.  Developments in ecological modeling.  Pages 37-53 in Agricultural
      nonpoint source pollution: modeling and application.  Elsevier, Amsterdam, The
      Netherlands.

Yereschukova, M. G., E.  Z. Hosseinipour, R. B. Ambrose, R. C. Russo, A. M. Nikanorov and
      V. V. Tsirkonov.  1992. A hydrodynamics and water quality modeling study of the
      Lower Don River,  Russia. Final Report 02.02-12, Environmental Research
      Laboratory, U.S. EPA, Athens, GA.
                                        76

-------
             EARLY DEVELOPMENT OF STRUCTURE-ACTIVITY
                  RELATIONSHIPS FOR ENVIRONMENTAL
                         TOXICOLOGY IN RUSSIA

                     Robert L. Lipnlck1, Vladimir A. Fttov2, and
                              Rosemarte C. Russo3
                                 ABSTRACT

      Quantitative structure-activity relationships have proved to be an essential tool in
assessing the risks to humans and the environment from exposure to industrial organic
chemicals.  One of the pioneers in the application of structure-activity relationships to
toxicology was Nikolai Vasilyevich Lazarev (1895-1974) at the Leningrad Institute of
Industrial Hygiene and Worker Safety.  Among his accomplishments was the
recognition that toxicity could be limited by water solubility, a finding that has been
applied decades later by the U.S. Environmental Protection Agency in defining an
appropriate set of tests for a chemical under review. His pioneering work, then,
presaged modern approaches in drug design and the assessment of toxicological
effects to man and to organisms in the environment.
                                INTRODUCTION

      Of increasing international concern is our ability to Identify those untested
industrial organic chemicals likely to have the potential to produce adverse effects to
human health and to the environment.  Even with active international cooperation
and data sharing, resources are not likely to be adequate in the foreseeable future to
test all industrial chemicals for all possible effects.  For this reason, environmental
regulatory agencies need scientifically defensible methods for setting testing priorities.

      Ideally, one would like to be able to write the structure of an organic
compound and use this information to predict its spectrum of possible or likely
1U.S. Environmental Protection Agency, Office of Pollution Prevention and Toxics (TS-796),
      Washington, DC 20460, USA.

"Department of Chemistry, Biology and Toxicology of Anti-tumor Compounds, Petrov Research
      Institute of Oncology, Leningradskaya St. 68, Pesochny-2, St. Petersburg 189646, Russia.

3U.S. Environmental Protection Agency, Environmental Research Laboratory, College Station
      Road, Athens, GA 30613, USA.

                                  77

-------
lexicological effects.  Such approaches-referred to as quantitative structure-activity
relationships (QSAR)--have been used for some time in pharmacology, and are
becoming of increasing importance in toxicology.

      Under Section 5 of the Toxic Substances Control Act, the U.S. Environmental
Protection Agency is authorized to asssess, prior to their manufacture, what risks new
industrial chemicals may pose to human health and the environment.  Under this
legislation, however, the chemical manufacturer is not required to provide EPA with
any data other than that in its files at the time of the submission.  EPA scientists must
then determine if a case can be developed as to why testing needs to be performed
prior to introductbn of the chemical into the marketplace.  A parallel chemical
selection and review process exists for chemicals already in commercial production.
Structure-activity relationships have provided an essential tool to EPA in meeting these
statutory requirements, for the more than 20,000 cases thus far reviewed.

      In 1975 (WHO, 1979), an international symposium Methods Used In the USSR for
Establishing Biologically Safe Levels of Toxic Substances was held in Moscow,
sponsored by the World Health Organization.  Part of the Symposium was devoted to
estimations of toxfclty from physicochemical properties and chemical structure, which
made reference to the pioneering work of Nikolai Vasilyevich Lazarev (1895-1974) (Fig.
1) in Leningrad. Lazarev's contributions have been extensively cited in a Russian
toxicology textbook (available in English) by four of his students (Fitov, et al., 1979).
        Fig. 1  Portrait of Nikolai Vasilyevich Lazarev (courtesy of V.A. Fitov).
                                      78

-------
                         LA2AREVS PIONEERING STUDIES

DEVELOPMENT OF INDUSTRIAL HYGIENE STANDARDS

      Lazarev's application of structure-activity relationships to toxicology provided
the foundation for the development of industrial hygiene standards in the Soviet Union
(Lipnick and Fitov, 1992). White leader of a group of toxicologists at the "Krasny
Treugolnik" (Red Triangle), Lazarev became interested in developing generalizations
about the toxicity of simple nonetectrolytes based upon his work on petroleum
hydrocarbons (Lazarev, 1931). He continued this work at the Leningrad Institute of
Industrial Hygtene and Worker Safety, where in 1932 he began serving as head of the
toxicology laboratory. He and his co-workers studied the toxicity of over 200 industrial
chemicals at this Institute from 1941. During these studies, they also discovered
systematic relationships between the chemicals' potency and certain
physicochemical properties. These relationships were documented in a number of
books (Lazarev and Astakhantsev, 1933; Lazarev, 1938, 1940, 1941, 1944, and 1958).

LAZAREV'S DATABASE

      In an effort to investigate the Interrelationship of various physicochemical
properties such as water solubility, partition coefficient, boiling point, and molar
refractivity to one another, Lazarev assembled an extensive set of data from the
literature for his analyses. In his 1944 monograph (Lazarev,  1944)  Nonelectrolytes,
subtitled Quantitative Classification of Biological, Physical, and Chemical Properties
(Fig. 2),  he used these data to develop a systematic approach for the classification of
nonelectrolytes.

GROUP  NUMBER

      Lazarev confirmed the earlier discovery made independently by Meyer (Lipnick,
1989a) and Overton (Lipnick, 1986; Overton, 1991) that the partition coefficient
between water and an immiscible organic phase provides the most useful property for
describing the biological activity of simple nonetectrolytes. For the sake of simplicity,
Lazarev proposed the use of a classification system involving nine groups. He chose
the olive oil/water partition coefficient as the basis for his physicochemical
classification of nonelectrolytas. Compounds belonging to group 1  (the most
hydrophilic compounds), were termed weak nonelectrolytes. This group includes
compounds for which log K,,,/^., falls within the range of -3 to -2. In Lazarev's group
number system, the most hydrophobic compounds were classified as group 9. These
were termed  strong nonelectrolytes and  have log «<„,„,«„,„ values within  the range 5 to
6.

      Fig. 3 shows a plot of the relationship between Lazarev's Group Number and
chain length for nine different homologous series of compounds. Lazarev found this
type of systematic behavior very useful for predicting the Group  Number for
compounds for which measured values were not available, and which is analogous to
the re- and fragment constant systems for estimating partition coefficient from
chemical structure (Hansch and Leo, 1979; Rekker, 1977; Rekker and Mannhold,  1992).
Once he was able to establish relationships between Group Number and biological

                                      79

-------
               MEAHKO-CAHHTAPHOE  VtlPABJlEHHE  BM<&

               BOEHHOMOPCKAH MEflHUHHCKAH AKAflEMHH



                           nojIKOBHHK MCA. CJiyivfibl
                          npo<|).  H. B.  JIA3APEB
                    H E 9 JI E K T P 0 JI H T bl
                  OFlblT EMO;iOrO-H3HKO-XMMMMECKOfi
                            MX CHCTEMATHKH
                                H 3 A A H H E
                       BOEIIHO-MOPCKOA MEAHUHHCKOfl AKAAEMHH
                                HEHHHTPHH 1*44
                                          *

      Fig. 2 Title page of Neelektrollty (courtesy of the Library of Congress).

transport and response, this method of estimating partition coefficients became
critical for the application of predictive methods in toxicology.

USE OF A LOG-LOG SCALE

      Lazarev investigated relationships between partition coefficient and biological
activity graphically using log-log plots.  Such plots provided the means to compare
data over nine orders of magnitude, and presaged the use of logarithmic
transformations in the development of extrathermodynamic linear free energy
relationships (Leffler and GrunwakJ, 1963) and quantitative structure-activity
relationships (Hansch and Leo, 1979).


                                      80

-------
       CD

       O
       CO
       E
       o
       c
       o
       u
       <« —
       o
       t_
       CD
       .O
       E
       3
                                        KOIL/WATER
                    123456789
                                      Group

Fig. 3 Relationship between Lazarev's Group Number and chain Jength for nine
      different homologous series of compounds (adapted from Fig. 4, Lazarev, 1944).
SOLUBILITY IN WATER

      A typical plot from Neelektrollty is shown in Fig. 4, depicting the relationship
between water solubility and oil/water partition coefficient. The data have been
presented as partition coefficient ranges defined by Lazarev's Group Number with the
lowest and highest water solu-bilities within each of these Groups represented by
open and closed circles, respectively. Fig. 5 is a  similar plot, with data on compounds
whose molecular weight is less than 100 Daltons.  In this case, the correlation is much
better4.  In contrast, Lazarev found that the relationship between group number and
log Kwcrt
-------
CORRELATION OF GROUP NUMBER WITH TOXICITY

      Fig. 6 shows a log-log plot of ocular irritation and tongue irritation versus group
number.  In each case there is an increase (i.e., lower concentrations of test
substance are required to produce a standard toxlcotogical response) in irritant action
with increasing group number.  Similarly, Fig. 7 presents a log-log plot of hypnotic dose
of urea derivatives as a function of log
      Lazarev was able to demonstrate similar graphical relationships between
partition coefficient or water solubility for various other effects such as arrest of
Isolated frog heart, contraction in isolated segments of frog heart ventricle, paralysis of
isolated rabbit intestine, narcosis in tadpoles and small fishes, and concentration in
mammalian blood associated with changes in reflex time, narcosis, respiratory failure,
or death.

EFFECT OF METABOLISM ON ELIMINATION

      Lazarev recognized that an important function  of metabolism was to Increase
the rate of elimination of foreign substances by decreasing their lipophlUcity, as
               |2

               ^^*
               cn 1
               CT>
               O
               -•o

                -1


                -2


                -3
                         COMPLETE  MISCIBILITY
                   -421
                                         50
                                          oAVERTIN
                      I     I	I	I	I
                                   0    12345

                                    K
                     -3   -2
                                 I nn k
                                      OIL/WATER
                        1    23456789

                                    Group

Fig. 4  Relationship between water solubility and oil/water partition coefficient
      (adapted from Fig. 10, Lazarev, 1944).
                                      82

-------
               E
               en
               o
                 0

                -1

                -2

                -3
                         MI5CIBLE
                -045"
                                 '-048
                      -072
           I    I    I    I
-3-2
      0
       K
        OIL/WATER
23456

        Group
                                            2    34   5


                                                  789
Fig. 5 Relationship between water solubility and oil/water partition coefficient for
      compounds with molecular weight less than 100 Daltons (adapted from Fig. 12,
      Lazarev. 1944).
illustrated In Fig. 8.  Thus, benzene upon its biotransformation to phenol and then to
hydroquinone changes from group 6 to group 2.  Similarly, metabolism of ethytene
glycol to oxalic acid produces a change from group 2 to group 1.

MATHEMATICAL PREDICTIVE MODELS

      From his data base and log-log plots, Lazarev derived mathematical models
relating partition coefficient to physteochemteal properties. In vttro effects, and In vivo
effects.

      Physlcochemical propwttot

      to9 Kb^^/wo*. = 1 -06 tog K01/walw + 0.32

      tog S = -0.89 tog K,,,/,,^^- 3.15, where S = Solubility in water (mmol/L).

      In VHm effects

      tog Ch9mot^ = -0.62 log K^/^,^, + 2.45,  where C = Effect concentration (mmol/L).
                                      83

-------
      In Vivo effects
           tadpoi, narco*
               -°-69 log
                                    + 1 .06. where C = Effect concentration
(mmol/L).

      tofl ctadpoie norco* = °-72 tog S - 1 .04, where C = Effect concentration (mmol/L).

      !°9 Ccc****** = °-72 toQ s - O-60' where C =. Effect concentration (mmol/L).
      Although these models were apparently not derived using regression analysis,
they serve the same function as that served today by certain statistically-derived
QSAR models in predictive toxicology (Lipnick 1990).

WATER SOLUBILITY CUTOFF

      Lazarev, like Overton, recognized that toxicity could be limited by water
solubility. Lazarev appears to be the first to have represented this phenomenon
graphically, and to have predicted the cutoff point by solving the two tog-log
4


3
     §2
     E
       0
                        3,89
                                 3.61
                                       3.13
              2.0
Tongue Irritation

   .62
                           1.180
                          Chloral
                                 0.70°
                                 Phenol
                                       0.98

                                          Ocular Irritation
                                                    .30
                                                      20
               -3    -2
                   \
                     -1012
                        L°9 K OIL/WATER
                                                   7
                     8
                         23456
                                   Group
Fig. 6  Relationship between Lazarev's Group Number and ocular and tongue irritation
      (adapted from Fig. 57, Lazarev, 1944).

                                     84

-------
functions of partition coefficient.5 A graph showing this cutoff for In vitro hemolysis is
depicted in Fig. 9.
     USE OF THE WATER SOLUiUJTY CUTOFF IN EPA ENVIRONMENTAL ASSESSMENTS

      The solubility-based cutoff in toxicity discussed above has been applied in a
practical way in defining  an appropriate set of tests for a chemical under review. For
example, for simple nonetectrolytes acting by a non-specific mechanism such as the
ones Lazarev used in his models, the concentration C (moles/L) producing narcosis in
the tadpole Rana temporarla can be described by the following QSAR model
(Lipntek, 1989b), expressed as the Hansch convention (reciprocal of biological
activity):

      tog (1/C) = 0.909 tog P + 0.727, where P = octanol/water partition coefficient.

      As an example of the use of this relationship and the solubility cutoff criteria
discussed previously, experimental toxicity tests show that phenanthrene produces
tadpole narcosis at 2 mg/L, which compares well with the QSAR prediction of 2.78
mg/L. Thus, for phenanthrene, the toxicity effect occurs close to, but does not
exceed, the water solubility (Lipntek, 1989b):

      Log P (octanol/water)6 = 4.49
      Melting point * 97.5 to 98.5°C
      Water solubility = 3.3 mg/L
      Molecular weight = 178.22

Similarly, Lazarev's oil-water correlation predicts effects at 1.5 mg/L; his solubility
correlation at 0.9 mg/L. Therefore, no soRjblHty cutoff for phenanthrene should have
been expected.

      On the other hand, anthraquinone shows no effect at saturation for the same
test. This should not be surprising since with a melting point of 286° and  log P
(octanol/water) value of 2.72, the predicted water solubility is 0.051 mg/L, which is
greatly exceeded by the predicted toxic concentration from the above QSAR of 132
mg/L.

      Log P (octanol/water) = 2.72
      Melting point = 286°C
      Water solubility (est.) = 0.051 mg/L
      Molecular weight = 208.22

Therefore, the solubility  cutoff should be anticipated for this chemical.
5The position of this cutoff would shift to lower group numbers as a function of increasing melting
      point for liquid solutes (Ljpnfck, 1990).

"Calculated using the CLOGP computer program, Version 3.3 (Leo and Weininger, 1985)
                                       85

-------
                                 CONCLUSION

      Nikolai Vasilyevich Lazarev was clearly a scientist ahead of his time. The
methods that he pioneered were not widely known outside of Russia, but presaged
modern approaches in drug design and the assessment of toxlcotogical effects to
man and to organisms in the environment.
                                  REFERENCES

Banerjee, S., S.H. Yalkowsky, and S.S. Valvani. 1980. Water solubility and octanol/water
      partition coefficients of organics. Limitations of the solubility-partition coefficient
      correlations, Environ. Sci. Technol., 14, 1227-1229.

Filov, V.A., A.A. Golubev, E.I. Liublina, and N.A. Tolokontsev. 1979. Quantitative
      Toxicology. John Wiley, New York.

Hansch, C. and A. Leo. 1979.  Substituent Constants for Correlation Analysis in
      Chemistry and Biology.  Witey-lnterscience, New York.
            en
                °5
             en
             o
            O
             u
             o
             c.
             a.
              "
0
               -0.5
                     -3
             -2
0
                                   LogK
                                        'HEXANE/WATER
 Fig. 7  Relationship between intraperitoneal hypnotic dose of urea derivatives to white
       mice and hexane/water partition coefficient (adapted from Fig. 74, Lazarev,
       1944).
                                       86

-------
           -3
                                                       8
Fig, 8 Effect of metabolism on decreasing lipophilicity for benzene, toluene,
      ethylbenzene, xytene, methanol, and ethytene glycol (adapted from Fig. 75,
      Lazarev, 1944).
Lazarev, N.V.  1931. Benzin kak Promyshtennyi lad (Benzene as an Industrial Toxicant),
      Sotsekgiz.

Lazarev, N.V, and P.I, Astakhantsev. 1933. Khimicheski Vrednye Veshchestva v
      Promyshtennosti. 1. Organlcheskie Veshchestra (Harmful Chemical Substances in
      Industry. 1. Organic Substances), Khkntekhizdat.

Lazarev, N.V.  1938.  Obshchiye Osnovy Promysgkebbiu Tokslkotogii (General Principles
      of Industrial Toxicology), Medgiz.

Lazarev, N.V,  1940.  Narkotiki (Narcotics), Institut Gigiyeny Truda I Profzabotevaniy.

Lazarev, N.V.  1941.  Biologischeskoye Deistviye Gazov pod Davleniyem (Biological
      Action of Gases Under Pressure).
                                      87

-------
lazarev, N.V. 1944.  Neetekrolity (Nonetectrolytes). Voennomorskaya, Medicinskaya
Akademiya.

Lazarev, N.V. 1958.  Obshchee Ucheniye o Narkotikakh i Narkoze (General Theory of
      Narcotics and Narcosis) Voenno-Morskaya Meditsinskaya Akaderniya.

Leffler, J.E. and E. GrunwakJ.  1963. Rates and Equilibria of Organic Reactions  John
      Wiley, New York.

Leo, A. and D. Weininger. 1985. Medchem Software Release 3.33, Medicinal
      Chemistry Project, Pomona  College, Ctaremont, CA.

Lipntek, R.L.  1986.  Charles Ernest Overton. Narcosis Studies and a Contribution to
      General Pharmacology.  Trends Pharmacol. Sci., 5, 161-164.

Lipnfck, R.L.  1989a. Hans Horst Meyer and the LipokJ Theory of Narcosis. Trends
      Pharmacol. Scl., 10, 265-269.
            o
            en
               o
             -2
                 _  -4.93
                            104
                     -2
0
   2    3
OIL/WATER
 567
                                                       8
                                   Group
Fig. 9  Relationship between water solubility (solid circles and solid line) and In vitro
      hemolysis (open circles and dashed line) as a function of oil/water partition
      coefficient. The point of intersection represents the highest partition coefficient
      or lowest water solubility at which hemolysis is not water solubility limited
      (adapted from Fig. 64, kazarev, 1944).
                                     88

-------
Lipnick, R.L. 1989b. A quantitative structure-activity relationship study of Overton's
      data on the narcosis and toxicity of organic compounds to the tadpole, Rana
      temporarla, in Aquatic Toxicology and Environmental Fate: Eleventh Volume
      (G.W. Suter, II, and M.A. Lewis, Eds.), ASTM STP 1007, American Society for
      Testing and Materials, Philadelphia, pp. 468-489.

Lipnick, R.L. 1990. Narcosis: fundamental and baseline toxicity mechanism for
      nonelectrolyte organic chemicals.  In W. Karcher and J. Devillers (Eds.) Practical
      Applicatbns of Quantitative Structure-Activity Relationships (QSAR) in
      Environmental Chemistry and Toxicology. Kluwer, Dordrecht.

Lipnick, R.L. and V.A. Fitov.  1992.  Nikolai Vasilyevich Lazarev, toxteobgist and
      pharmacologist, comes in from the cold. Trends Pharmacol. Sci., 13, 56-60.

Overton, C.E.  1991.  Studies of Narcosis.  (R.L. Lipnick, Ed.), Chapman and Hall,
      London.

Rekker, R.F. 1977. The Hydrophobic Fragmental Constant.  Ebevier, Amsterdam.

Rekker, R.F. and R. Mannhold.  1992.  Calculation of Drug Lipophilicity. VCH,
Weinheim.

WHO, World Health Organization. 1975.  Methods Used in the USSR for Establishing
      Biologically Safe Levels of Toxic Substances.  World Health Organization,
      Geneva, Switzerland.
                                     89

-------
                MERCURY CONTENT AND ULTRASTRUCTURE OF GILLS
        AND SCALES OF FISH FROM LAKES IN NORTH AND NORTHWESTERN
            RUSSIA THAT ARE POLLUTED BY ATMOSPHERIC DEPOSITION
               Terry Haines", Victor Komovb, Charles Jagoec, and Victoria Mateyb
                                        ABSTRACT

       Field studies were conducted of lakes in Darwin National Reserve in 1989, and in Karelia
in 1991.  Darwin National Reserve is 300 km north of Moscow at the west end of Rybinsk
Reservoir.  The terrain is generally flat and consists of thick sandy till covered with hardwood
forest. Karelia is north of St. Petersburg and consists of hilly terrain underlain with precambrian
bedrock.  In each case, remote lakes were visited and sampled for water and fish, primarily perch
Perca fluviatilis. Water samples were analyzed for pH, color, specific conductance, and major
cations and anions. Fish were weighed and measured, dorsal muscle tissue was collected and
analyzed for mercury, and gills and scales were sampled for examination by light and electron
microscopy. In both locations acidic lakes (acid neutralizing capacity <0) were common. Acidic
lakes were both clear and colored and the dominant anion in both types was sulfate, indicating
that the lakes were acidic because of atmospheric deposition of strong acids and not because of
organic acids.  Mercury content of fish was increased in acidic and in colored lakes.  Mercury
appears to enter the lakes by atmospheric deposition, as there are no local  sources.  Organic acids
are believed to increase mercury bioavailability in lakes by transporting mercury from terrestrial
regions and possibly contributing to methylation. Also, mercury methylation is probably enhanced
in acidic  lakes, increasing bioavailability. Gills of perch from acidic lakes had thicker lamellar
epithelia and more ion-transporting cells than those collected in circumneutral lakes.  The density
of microridges on gill surfaces was reduced in perch from acidic lakes. These differences in gill
structure serve to decrease gill surface area, increase diffusion distances, increase active ionic
influx, and represent responses that allow perch to inhabit acidic waters.  Scales of perch from
acidic lakes had little or none of the pattern of ridges at  the foci observed in scales of perch from
higher pH  lakes.  This probably results from development and growth in low pH, low Ca water,
and may be a  useful indicator of stressful acidic conditions during early life stages.
"U.S. Fish and Wildlife Service, National Fisheries Contaminant Research Center, Field Research
Station, 313 Murray Hall, University of Maine, Orono, Maine 04469 USA
Institute for Biology of Inland Waters, Academy of Sciences of Russia, 152742 Borok, Nekouz,
Yaroslavl. RUSSIA
"University of Georgia, Savannah River Ecology Laboratory, P.O. Drawer E, Aiken, South Carolina
29802 USA
                                              90

-------
                                      INTRODUCTION

       Atmospheric transport and deposition of contaminants is a world-wide problem.  The
importance of atmospheric transport in contamination of aquatic ecosystems was first documented
for emission of acidic gases such as sulfur oxides and nitrogen oxides and the deposition of strong
acids, which resulted in water acidification and loss of fish populations (Schindler 1988).  This
process is also important in the mercury contamination of fish in waters of low acid neutralizing
capacity (ANC) in northern Europe and North America  (Lindqvist et al. 1991; Spry and Wiener
1991).  Further, lake acidification increases mercury concentration in fish (Grieb et al. 1990;
Hakanson et al. 1990; McMurtry et al. 1989).
       Atmospheric deposition of sulfur compounds in Europe has been estimated using a model
developed by the Cooperative Programme for Monitoring and Evaluation of the Long Range
Transmission of Air Pollutants in Europe (EMEP).  The results of this model give a deposition rate
of 800-1,640 eq S/ha per year for northern Russia, which is comparable to or exceeds deposition
rates for regions in Scandinavia where lake acidification and fish mercury accumulation has
occurred (Chadwick and Kuylenstierna 1991).  In undertaking this study, our objectives were to
confirm the previously reported presence of clear-water and brown-water acidic lakes in two areas
of Russia to determine whether fish inhabiting these  lakes contained elevated levels of mercury,
and to determine whether these fish exhibited symptoms of damage from acidity.

                                         METHODS

       This study was conducted in two phases. In May and June 1989 we sampled lakes in
Darwin National Reserve, Yaroslavl Region, and in June 1991 we sampled  lakes in the Karelia
Autonomous Republic (Figure 1).  Darwin National Reserve is a protected natural area created in
1943, located at the western end of Rybinsk Reservoir.  It is characterized by thick tills and sandy
soils (Ager 1980), and the lakes are primarily precipitation-dominated seepage lakes. In Karelia
we visited lakes in the vicinity of Suoyarvi, which were primarily highly colored lakes, and lakes in
the vicinity of Kondopoga, which were primarily clear-water lakes. Both areas were underlain by
granitic bedrock covered with thin soil (Ager 1980), and the lakes were primarily drainage lakes.
Most lakes were in forested catchments with no dwellings or roads. An exception was Suoyarvi
Lake, which has a settlement at the southern end.
       At each lake, a water sample was collected by immersion of cleaned polyethylene bottles
just below the surface near the point of maximum depth.  Bottles were tightly capped, held in the
dark, and within a few hours analyzed for pH (Orion model SA210 meter equipped with a Ross
combination electrode), ANC (acid neutralizing capacity, measured by inflection point titration),
and true color (visual comparison with platinum-cobalt  standards).  The remaining water sample
from each lake was refrigerated and later analyzed for major cations and anions.  Calcium,
magnesium, and sodium were determined by flame atomic absorption spectropho tome try, chloride,
fluoride, sulfate, and nitrate by ion chromatography,  and total aluminum by graphite furnace
atomic absorption spectrophotometry.
       Perch (Perca fluviatilis) were collected by angling from various points around the shoreline
of each lake,  except for Suoyarvi Lake, where fish were captured in a gill net set overnight.  The
second and third gill arches were dissected from each fish and fixed in either 0.1 M cacodylate
buffer, pH 7.3, containing 2.5% glutaraldehyde, or in 0.1 M HEPES buffer,  pH 7.4, containing 1%
glutaraldehyde, 4% formaldehyde, and 5% sucrose. Fish were then placed in plastic bags and
frozen (-5°C)  within a few hours, and remained frozen until they were analyzed for mercury.  Fish
and fixed gill tissue were transported to the Institute for Biology of Inland  Waters, Borok, in 1989,
and the Institute for Research on Northern Fishes, Petrozavodsk, in  1991.
       At the laboratories, the fish were thawed, weighed to the nearest gram, and measured
(total length) to the nearest millimeter.  Scales were removed from the left side of the fish adjacent
to the dorsal  fin and placed in paper envelopes for future ultrastructure determination.  Fish were
                                              91

-------
                                                    T3

                                                    3
                                                    O

                                                    S
                                                    O
                                                    c
                                                    o>

                                                    2
                                                    to
                                                   I
                                                   IJJ
                                                   •s
                                                   Q.
92

-------
placed on an acid-washed plexiglass plate and a 2- to 3-g sample of skeletal muscle was dissected
from the left side of the fish, adjacent to the dorsal fin and extending from the rib cage to the
center of the dorsal surface. All dissecting instruments were stainless steel, and all instruments
and glassware were cleaned with 10% nitric acid and rinsed with distilled water.
       Moisture content of muscle tissue was determined by placing a portion of the tissue in a
dried, tared glass dish and weighing before and after drying at 105°C. The remaining tissue,
typically 1 to 2 g, was placed in a tared glass beaker, weighed, and digested in a 1:1 mixture of
nitric acid and hydrogen peroxide (FAO/SIDA 1983).
       Fish were processed in batches of 5 to 10.  Two fish from each batch were randomly
selected for use in determination of accuracy and precision of mercury determination. For these
fish, the tissue was divided into two nearly equal portions. A known amount of mercury standard
solution was added to one portion of tissue, and all portions were digested and analyzed.  A sample
of National Institute of Science and Technology (NIST) tuna reference material and a reagent
blank was included with each batch of samples.  After digestion, solutions were cooled and diluted
to 25 ml with distilled water, and 1- to 2-ml portions were analyzed for total mercury with a gold
film analyzer (Arizona Instrument Co., Jerome, Arizona USA).  Each digestate was analyzed in
duplicate. Mercury in all samples exceeded our calculated detection limit of 5 ng. The recovery of
mercury from spiked  samples averaged 95.4% (range 90-102%). The percent difference between
replicate samples averaged 2.9% (range 1.7-5.2%).  The measured mercury content of NIST tuna
averaged 0.93 /Ag/g (range 0.88-0.96 /ig/g), and all analyses were within the certified range for this
material.
       Gill  tissue was postfixed in buffered OsO4 for two hours and dehydrated through a graded
ethanol series into absolute acetone. Whole gill arches were critical point dried under liquid CO2,
sputter-coated  with gold, and examined using a JEOL JSM-25S electron microscope operated at 15
KV. Scales were first placed in glass vials with 5% NaOH for 6-12 h, then hydrolyzed in a
saturated solution of  K2Cr2O1 in 10% NaOH for 1-7 d. The scales were then washed  in distilled
water until all yellow color was removed, placed in 3:1 mixture of absolute ethanol and acetone
and then in absolute  acetone for 10-15 min each, air dried, mounted on stubs, and coated with gold
and examined in the  same manner as were gills.

                                          RESULTS

       The lakes surveyed varied in pH  from 4.6 to 8.1  and in color from 3 to 188 Hazen (Figure
2, Table 1).  Drainage and  seepage lakes were relatively uniformly distributed over the range of
color values, but all seepage lakes surveyed were acidic (pH <5.0). The lakes in Darwin Reserve
were predominately seepage lakes whereas the lakes in  Karelia were predominately drainage
lakes.
       The fish collected ranged in weight from 8 to 515 g, and in mercury content from 0.06 to
3.04 /ig/g wet weight  (0.76-17.9 pg/g dry weight) (Table 2). Omitting two unusually large fish from
Tyomnoye Lake reduces the maximum weight to 196 g and the maximum mercury content to 1.03
/ig/g wet weight (6.87 /*g/g dry  weight). Fish mercury content was correlated with fish size,
especially weight (R2=0.32, p=0.0001); therefore the least square mean mercury concentration,
which represents fish mercury  concentration normalized by fish weight, was computed by analysis
of covariance using weight as the covariate.
       Plotting fish mercury concentration against lake pH and color (Figure 3) indicates that fish
from high pH lakes have relatively low mercury content regardless of lake color but in low pH
lakes fish mercury content increases with lake color. Accordingly, stepwise regression analysis was
performed using the non-intercorrelated variables  pH, color, and sulfate,  with the data stratified
by lake drainage type.  The results (Table 3) indicate that color is more important in the
regression than is pH for seepage lakes, and that the relation between fish mercury and lake pH is
not significant in the  absence of color data for drainage lakes.  The inclusion of lake sulfate did not
significantly improve  the regression for seepage lakes.
                                              93

-------
m ®

13


§|
c Q.
2 $
Q CO

D O
0)
                         D
                       D  OD
                        O
                       ~
                       O
                        O
co
                 
-------
   TABLE 1.  PHYSICAL AND CHEMICAL CHARACTERISTICS OF THE STUDY LAKES.
Lakes
Darwin Lakes
Dorojiv
Dubrovskoye
Hotavets
Motykino
Rybinsk
Tyomnoye
Uteshkovo
Karelia Lakes
Blue Lamba
Chuchyarvi
Grushna Lamba
Ilyakalkenyarvi
Kabozero
Lamba Vegarous
Leukunyarvi
Sargozero
Suoyarvi
Uros
Vegarousyarvi
Venderskoye
Vuontelenyarvi
Mean Values
Drainage Lakes"
Seepage Lakes
Area
(ha)

200
20
160
2
395,000
20
5

307
112
3
104
210
7
-
200
6,070
426
1,880
998
394

1,045
84
Depth
(m)

3
2
3
4
30
2
3

4
5
5
6
3
4
-
4
5
3
5
6
3

4.2
3.5
Drainage
Type

Seepage
Seepage
Drainage
Seepage
Impoundment
Seepage
Seepage

Seepage
Seepage
Seepage
Drainage
Drainage
Drainage
Drainage
Drainage
Drainage
Drainage
Drainage
Drainage
Drainage

Drainage
Seepage
Color
(Hazen)

13
182
188
19
55
70
150

3
8
3
171
141
182
182
25
104
9
105
23
186

120
56
PH

4.6
4.6
8.1
4.8
8.0
4.7
4.8

4.6
5.0
4.6
4.6
5.5
4.5
4.9
7.9
5.8
5.9
5.1
7.0
4.6

5.8
4.7
ANC
(/teq/1)

-41
-50
227
-38
1,369
-53
-33

-17
-9
-19
-25
13
-39
-9
252
23
16
-2
167
-24

56
-33
S04
Oieq/1)

71
64
32
49
420
52
29

67
48
48
54
60
75
68
29
83
72
49
44
68

58
53
'Rybinsk omitted.
                                   95

-------
TABLE 2. MEAN AND RANGE OF FISH WEIGHT, MOISTURE CONTENT, AND MERCURY
CONCENTRATION IN FISH, AND LEAST SQUARE MEAN MERCURY CONCENTRATION IN
                     FISH FROM THE STUDY LAKES.
Lake
Blue Lamba
Chuchyarvi
Dorojiv
Dubrovskoye
Grushna Lamba
Hotavets
Ilyakalkenyarvi
Kabozero
Lamba Vegarous
Leukunyarvi
Motykino
Rybinsk
Sargozero
Suoyarvi
Tyomnoye
Weight, g
Mean
(Range)
22
(18-28)
59
(18-133)
62
(31-94)
27
(23-30)
38
(28-48)
62
(37-150)
35
(20-78)
35
(24-47)
55
(28-80)
30
(8-54)
60
(43-75)
86
(59-163)
18
(8-27)
98
(53-131)
155
(50-515)
Moisture, %
Mean
(Range)
85.6
(81.1-89.7)
90.6
(83.2-93.9)
80.9
(78.0-86.2)
83.6
(77.8-88.1)
81.5
(76.9-86.0)
88.6
(80.4-93.8)
82.4
(72.0-94.2)
81.5
(61.7-87.2)
80.5
(75.9-85.1)
88.2
(82.8-96.3)
81.4
(79.3-87.9)
80.8
(78.1-82.6)
85.2
(75.0-89.6)
81.5
(79.6-83.6)
79.6
(74.6-83.4)
Mercury, ju.g/g wet weight
Arithmetic Least Square
Mean Mean
(Range)
0.34
(0.26-0.44)
0.10
(0.08-0.13)
0.50
(0.36-0.71)
0.64
(0.56-0.68)
0.30
(0.20-0.43)
0.11
(0.08-0.16)
0.28
(0.19-0.36)
0.31
(0.22-0.39)
0.40
(0.30-0.52)
0.21
(0.17-0.24)
0.57
(0.43-0.97)
0.19
(0.08-0.45)
0.12
(0.09-0.18)
0.29
(0.19-0.38)
1.06
(0.45-3.04)
0.41
0.08
0.47
0.71
0.33
0.09
0.31
0.34
0.39
0.25
0.54
0.11
0.19
0.18
0.56
                                 96

-------
Uros
Uteshkovo
Vegarousyarvi
Venderekoye
Vuontelenyarvi
19
(14-21)
45
(18-98)
26
(17-45)
41
(12-107)
43
(16-196)
83.7
(76.1-90.3)
80.4
(75.7-85.8)
79.5
(76.9-82.8)
79.8
(72.9-81.0)
84.5
(76.9-93.0)
0.12
(0.06-0.18)
0.78
(0.54-0.97)
0.34
(0.20-0.44)
0.15
(0.10-0.24)
0.53
(0.32-1.03)
0.19
0.80
0.41
0.17
0.54
      TABLE 3.  STATISTICALLY SIGNIFICANT (P <0.05) REGRESSION EQUATIONS,
   CORRELATION COEFFICIENTS, AND PROBABILITY FOR STEPWISE REGRESSION OF
 LAKE PHYSICAL AND CHEMICAL VARIABLES ON FISH MERCURY CONTENT. ANALYSES
       ARE PRESENTED FOR ALL LAKES, AND STRATIFIED BY DRAINAGE TYPE.
Lake
Type
All


Seepage

Drainage

Number of
Variables
1
2
3
1
2
t
2
3
Equation
-0.0347 PH + 0.317
-0.0336 pH + 0.000144 color + 0.298
-0.0331 pH + 0.000143 color
-0.000019 sulfate + 0.297
0.000711 color + 0.128
0.000671 color - 0.173 pH + 0.946
-0.0260 pH + 0.000030 color + 0.252
-0.0256 pH + 0.0000284 color
-0.0000167 sulfate + 0.251
R2
0.46
0.49
0.50
0.56
0.69
0.68
0.68
P
0.001
0.0035
0.0119
0.0332
0.05
0.0058
0.0216
      The structure of the gills of perch has been previously described by Matey (1984). Primary
lamellae or filaments project in two parallel rows from the gill arch, and each primary lamella
supports two rows of secondary lamellae (Figure 4a). The epithelium covering the gill consists
primarily of thin, squamous respiratory cells, along with mucous cells and chloride (ion
transporting) cells, underlain by undifferentiated cells. Most chloride cells occur in the primary
lamellar epithelium, especially in the spaces between secondary lamellae. Normally, the surfaces
                                         97

-------
                                                   &
                                                   Q)

                                                   O

                                                   8
                                                   CL
                                                  JO

                                                  V)


                                                  2

                                                  5
                                                   8
                                                   §
                                                   £
                                                   CO


                                                   I
                                                   O

                                                   w
J8M 6/6ry 'UOJ1BJ1U90UOQ AjnOJ8|/\|
I
O)
                    98

-------
Figure 4.  Scanning electron micrographs of normal gill structure. A. Gill filaments with
secondary lamellae.  Bar= 1 mm.  B. Epithelial surface.  Bar= 10 /im.

of the respiratory cells are covered by a well-developed system of microridges, which may serve to
anchor mucus, increase surface area, or create micro-turbulence (Figure 4b).  This structural
arrangement is found in most teleost fishes (Hughes 1984; Laurent 1984).
       For the Darwin Reserve lakes, all fish examined from Hotavets Lake (pH 8.1) had normal,
well-formed gills. The primary lamellar epithelia contained  pavement cells with well-formed
microridges and few chloride cells (Figure 5a), and secondary lamellae were thin and regular
(Figure 5b). Perch from the more acidic lakes (Dubrovskoye, Dorojiv, Motykino, and Uteshkovo
lakes) all had alterations in gill structure. Chloride cell numbers were greatly increased along the
primary lamellae (Figure 5c).  Some chloride cells had apical crypts (Figure 5d), which were not
observed in fish from the higher pH lake.  Chloride cells were also present on many secondary
lamellae (Figure 5e).  Mucus secretion was elevated, as evidenced by increased numbers of mucus
droplets and secretory pores associated with mucous cell activity (Figure 5f).  The density of
microridges on the gill epithelia was reduced in fish from the acid lakes compared with fish from
the higher pH lake (Figure 5g;  compare with Figure 4b).
       For the highly colored lakes in Karelia, fish from lakes of pH <5.5 had thickened and
swollen secondary lamellae (Figure 6a). This phenomenon was especially severe in fkh from
Leukunyarve, Vegarousyarvi, and Lamba near Vegarous lakes, where  epithelial hyperplasia
produced regions of fused secondary lamellae  (Figure 6b). These fish also often had locally swollen
or evaginated regions on the  primary lamellae (Figure 6c), although these abnormalities were not
observed in fish from Ilyakalkenyarvi Lake, which had a pH of 4.6. Chloride cell number on
primary lamellae appeared to increase with decreasing pH, and was greatly elevated in fish from
Vuontelenyarvi, Ilyakalkenyarvi, and Lamba near Vegarous lakes (Figure 6d).  In perch from lakes
with pH below 5.1, chloride cells were also abundant on secondary lamellar epithelia (Figure 6e).
                                              99

-------
Figure 5.  Scanning electron micrographs of gills from Darwin Reserve fish. Bar=10 /Ltm in all
cases.  A. Primary lamellar epithelium of a fish from a high pH lake, showing respiratory
pavement cells and a lack of chloride cells.  B.  Secondary lamellar epithelium of a fish from the
same lake as A.  C. Primary lamellar epithelium of a fish from an acidic lake showing numerous
chloride cells (arrow).  D. Chloride cell with an apical crypt (arrow).  E. Secondary lamellar
epithelium of a fish from an acidic lake showing numerous chloride cells (arow). F. Secondary
lamellar epithelium of a fish from an acidic lake showing numerous secretory pores (arrow). G.
Primary lamellar epithelium of a fish from an acidic lake showing reduced microridge density.
                                             100

-------
Figure 6.  Scanning electron micrographs of gills of fish from colored, acidic Karelian lakes. A.
Swollen and thickened secondary lamellae (arrow). Bar=100 /xm. B. Epithelial hyperplasia
(arrow). Bar=l mm. C. Primary lamelar epithelial swelling (arrow). Bar=100 pm.  D. Primary
lamellar epithelium with numerous chloride cells (arrow). Bar=100 ^tm.  E. Secondary lamellar
epithelium with numerous chloride cells (arrow). Bar=10 p.
                                            101

-------
Figure 1.  Scanning electron micrographs of gills of fish from Karelian lakes.  A. Epithelial
hyperplasia leading to secondary lamellar fusion (arrow) in a fish from a colored, acidic lake.
Bar=100 ju.m.  B. Cell surface microridge density in a fish from a high pH lake.  Bar=10 /u,m. C.
Cell surface microridge density in a fish from a low pH, clear lake.  Bar=10 ju.m.  D. Cell surface
microridge density in a fish from a  low pH, colored lake.  Bar=10 ju,.
                                             102

-------
       Alterations in gill structure associated with low pH appeared less severe in fish from clear-
water lakes in Karelia than in those from colored-water lakes. Epithelial hyperplasia leading to
secondary lamellar fusions was observed only in fish from Chuchyarvi, pH 5.0 (Figure 7a).
Secondary lamellae were somewhat swollen and thickened in fish from the lower pH clear-water
lakes.  Chloride cell numbers were increased on both the primary and secondary lamellae in fish
from lakes with pH <5.9.  Compared with perch from lakes of higher pH (Figure 7b), cell surface
microridge density was slightly decreased in fish from the most acidic clear-water lakes (Figure
7c).  In contrast, fish from low pH, colored-water lakes had substantially reduced microridge
patterns, with the typical labyrinth appearance replaced by shorter, smaller surface ridges (Figure
7d).
       Perch have typical ctenoid scales (Figure 8a).  The scales of perch from Hotavets (pH 8.1)
were characterized by a focus consisting of a number of microridges creating a complex pattern
(Figure 8b). These microridges were lacking in the scales of the fish from lakes in Darwin Reserve
with pH 4.4-4.8 (Figures 8c, 8d).

                                        DISCUSSION

       The areas of Russia where we surveyed lakes receive relatively high levels of sulfur
deposition.  Karelia receives 420-800 eq/ha/yr, comparable to southern Finland,  Sweden, and
Norway, and Yaroslavl region receives 800-1640 eq/ha/yr, comparable to northern Germany
(Chadwick and Kuylenstierna 1991).  The critical loading of acidity is estimated at 200 eq/ha/yr or
less for Karelia and 200-500 eq/ha/yr for Yaroslavl region, and deposition is estimated to exceed
critical loading by 500-1000 eq/ha per year in Karelia, but only 200 eq/ha per year or less in
Yaroslavl region (Hettelingh et al. 1991). Consequently, we found no acidic drainage lakes in
Darwin Reserve.  The seepage lakes were precipitation-dominated as reflected in the low ion
content, and were not influenced by the chemistry of the soils or surficial material.  The acidity of
the seepage lakes in Darwin Reserve reflects the acidity of the precipitation they receive. In
Karelia, both drainage and seepage lakes were acidic. The situation here is similar to that in
Finland, Sweden, and southern Norway, where acidic precipitation falling on a landscape
consisting of resistant bedrock and thin, nutrient-poor soils has resulted in the acidification of
thousands of lakes, with resultant biological effects (Rosseland et al.  1986).
       The mercury content of perch from undisturbed lakes in Russia is comparable to that
reported for this species from similar lakes in Sweden and Finland, disregarding the two large fish
from Tyomnoye lake, which had unusually high mercury concentrations.  Mercury concentrations
were reported to be 0.2 to 2.0 /Ltg/g dry weight for "small" perch and 0.6 to 6.0 /ig/g dry weight for
"large" perch in Swedish forest lakes (Lindqvist et al. 1991).  Mercury content of perch muscle
tissue  in 11 Swedish lakes of pH 5.2-6.2 was between 0.04 and 0.29 pg/g wet weight (Paulsson and
Lundbergh 1991).  In Finland, mercury concentrations between 0.03-0.53 pg/g in perch from
circumneutral lakes and 0.15-0.63 in fish from acidic lakes have been noted  (Verta et al. 1990).
The larger fish from the most acidic lakes in our study approached or exceeded  mercury
concentrations believed harmful to fish consumers (Wiener 1987;  Grant 1991).
       The association of elevated mercury content in fish with low water pH has been widely
reported (Lindqvist et al. 1991; Spry and Wiener 1991).  Our results confirm these findings and
extend them to two regions of Russia not previously investigated. However, the phenomenon is
much more complex than a simple water acidity-fish mercury relation. An extensive study of
mercury accumulation by fish in  Swedish forest lakes led to the conclusion that concentrations
were determined by the bioavailability of mercury to lower trophic levels, and that lake humicity,
productivity, and acidity controlled bioavailability  (Lindqvist et al. 1991).  Our results suggest that
acidity and humicity are important factors in affecting fish mercury content in Russian lakes (we
did not investigate productivity). We found low concentrations of mercury in fish  from high pH
lakes regardless of color, which suggests that acidity is a necessary but not sufficient factor in
mercury accumulation by fish in  these lakes.
                                             103

-------
                                   S
104

-------
       Humic matter was found to control the solubility and watershed export of mercury
deposited in precipitation in Sweden and Canada (Iverfeldt and Johansson 1988; Mierle and
Ingram 1991).  Further, water concentrations of mercury were highly correlated with water color
(Meili et al. 1991). Most of the mercury in clearwater lakes was deposited in the sediment whereas
most of the mercury in humic lakes was retained in the water column (Meili 1991).  Thus humic
matter may affect fish mercury content by affecting the mercury loading to a lake and by retaining
mercury in the water column where it is available for uptake  in biota.
       Alterations in gill morphology have been reported in a number of fish species in response to
water acidification (Jagoe and Haines 1983; Matey 1984; Evans et al. 1988; Tietge et al. 1988).
Most studies have examined effects after laboratory exposure; relatively few studies have been
conducted on wild fish from chronically acidic environments.  Chevalier et al. (1985) reported gill
abnormalities associated with acidification in brook trout (Salvelinus fontinalis) collected from
acidic lakes, and Leino et al. (1987) found changes in the gill epithelium of two species of cyprinids
collected from an experimentally acidified lake. Our results expand these observations to perch
inhabiting acidified environments in Russia, and allow comparisons from colored and clear acidic
waters.
       Exposure to low pH water causes osmoregulatory and ionoregulatory disturbances in fish
(McDonald 1983), primarily because of increased passive efflux of sodium.  The changes in gill
morphology we observed may represent  adaptation of acclimation to minimize this effect.  The
thickening of the epithelium would lengthen the diffusion pathway by which ion losses occur,  and
increasing numbers of chloride cells would allow increased active uptake of ions from the
environment.  Perch are known to  be relatively acid tolerant and are frequently the only species of
fish inhabiting acidic lakes (Rask and Tuunainen 1990), including the ones we studied. Perhaps
these changes in gill morphology allow this species to survive at low pH.
       Ctenoid fish scales consist of two layers: an upper, ridged osseous layer with cteni and a
lower fibrillary plate (Fouda 1979). The focus is formed early in the life of the individual (Sire
1986) and does not change appreciably after formation. Thus the morphology of this structure
reflects environmental conditions early in the life of the fish.  The loss of microridges from the
scale focus in fish from acidic lakes may reflect disruption of calcium metabolism in these fish.
Steingraeber and Gingerich (1991) found that brook trout exposed to pH 5 water had reduced body
calcium, and Reader et al. (1989) found  a similar effect for brown trout (Salmo trutta) at pH 4.5.
Majewski et al. (1990) found reduced bone calcium concentration in adult Atlantic salmon (Salmo
solar) held in a river of pH 4.7-5.2. Therefore, scale structure may be a sensitive measure of
physiological stress from water acidity in fish. Because scale  structure remains unchanged after
formation, examination of scales of older fish could be used as an indicator of water chemistry at
the time the scale was formed.

                              SUMMARY AND CONCLUSIONS

       We surveyed 20 lakes in two regions of Russia to assess lake acidity, and to determine the
effects of water acidity on fish mercury content and gill and scale ultrastructure. The lakes
surveyed ranged in pH from 4.5 to 8.1 and in color from 3 to 188 Hazen. The mercury content of
fish ranged from 0.06 to 3.04 fig/g wet weight, which was comparable to that reported for this
species from forest lakes in Sweden and Finland. Inasmuch as these were all remote lakes with no
local sources of pollution, atmospheric deposition is presumed to be the source of both acidity  and
mercury. Lake acidity and color, or humic content,  were the major lake characteristics related to
fish mercury content. Fish from high pH lakes were low in mercury regardless  of other lake
characteristics. Fish from low pH  lakes varied widely in mercury content, with  fish from colored
lakes having higher concentrations than those from clear lakes. Regressions including lake pH
and color, separated by drainage type, explained about 70% of the variance in fish mercury
content.  Acidity and color may affect fish mercury content by regulating loading and
bioavailability of mercury to lower trophic levels in these lakes.
                                             105

-------
       Gills and scales of fish from acidic lakes were morphologically different from those of fish
from circumneutral lakes. In acidic lakes, fish gills had thickened secondary lamellae, increased
numbers of chloride cells, apical crypts in some chloride cells, increased mucous production, and
decreased microridge density.  These changes were more severe in highly colored acidic lakes than
in clear-water lakes. Scales of fish from acidic lakes lacked microridges in the focus.  The gill
abnormalities are all changes that would tend to reduce the effects of ion loss, which is the major
physiological effect of water acidity to fish.  The scale abnormality may result from disrupted
calcium metabolism in acidic lakes.

                                   ACKNOWLEDGMENTS

       Financial support for this study was provided by U.S. Fish and Wildlife Service, U.S.
Environmental Protection Agency, and U.S. Department of Energy through contract DE-AC09-
76SROO-819  to the University of Georgia's Savannah River Ecology Laboratory in the United
States, and the Institute for Biology of Inland Waters, Russian Academy of Sciences in Russia.
Research space and facilities was provided by the Institute for Research on Northern  Fishes in
Petrozavodsk, and by the Laboratory of Physiology and Toxicology, Institute for Biology of Inland
Waters in Borok.

                                       REFERENCES

Ager, D.V. 1980. The Geology of Europe.  McGraw-Hill Book Co., New York.

Chadwick, M., and J. Kuylenstierna. 1991.  The relative sensitivity of ecosystems in Europe to
       acidic depositions. A preliminary assessment of the sensitivity of aquatic and terrestrial
       ecosystems. Perspectives in Energy 1:71-93.

Chevalier, G., L. Gauthier, and G. Moreau.  1985.  Histopathological and electron microscopic
       studies of gills of brook trout, Salvelinus fontinalis, from acidified lakes.  Can. J. Zool.
       63:2062-2070.

Evans, R., S. Brown, and T. Hara. 1988.  The effects of aluminum and acid on gill morphology in
       rainbow trout, Salmo gairdneri. Environ. Biol. Fish. 22:299-311.

FAO/SIDA. 1983.  Manual of methods in aquatic environment research, Part 9, Analyses of metals
       and organochlorines in fish, FAO Fish. Tech. Pap.  212.

Fouda, M. 1979.  Studies on scale structure in the common goby Pomatoschistus micropa Kroyer.
       J. Fish Biol. 15:173-183.

Grant, L. 1991. Indirect health effects associated with acidic precipitation.  Acidic Deposition:
       State of Science and Technology, Summary Report of the U.S. National Acid Precip. Assess.
       Program, Chapter 23. P. Irving, ed. National Acid Precipitation Assessment Program,
       Washington.

Grieb, T., C.  Driscoll, S. Gloss, C. Schofield, G. Bowie, and D. Porcella. 1990. Factors affecting
       mercury accumulation in fish in the Upper Michigan Peninsula.  Environ. Toxicol. Chem.
       9:919-930.

Hakanson, L., T. Andersson, and A. Nilsson.  1990. Mercury in fish in Swedish lakes  - linkages to
       domestic and European sources of emission. Water Air Soil Pollut. 50:171-191.
                                              106

-------
Hettelingh, J-P., R. Howning, and P. de Smet. 1991. Mapping critical loads for Europe.
       Coordination Center for Effects, Tech. Rep. 1, Bilthoven, the Netherlands.

Hughes, G. 1984. General anatomy of the gills. In:  Fish Physiology, Vol. 10A, Gills: Anatomy, Gas
       Transfer, and Acid-Base Regulation, W. Hoar and D. Randall, eds.  Academic Press, New
       York, pp 1-72.

Iverfeldt, A. and K. Johansson. 1988. Mercury in run-off water from small watersheds.  Verh.
       Internat. Verein. Limnol. 23:1626-1632.

Jagoe, C., and T. Haines. 1983. Alterations in gill epithelial morphology of yearling Sunapee trout
       exposed to acute acid stress. Trans. Am.  Fish.  Soc. 112:689-695.

Laurent, P. 1984. Gill internal morphology.  In: Fish Physiology, Vol. 10A, Gills: Anatomy, Gas
       Transfer, and Acid-Base Regulation, W. Hoar and D. Randall, eds.  Academic Press, New
       York, pp 73-183.

Leino, R., P. Wilkinson, and J. Anderson. 1987.  Histopathological changes in the gills of pearl dace
       Semotilus margarita and fathead minnow Pimephales promelas from experimentally
       acidified Canadian lakes.  Can. J. Fish. Aquat. Sci. 44 (Suppl. 1): 126-134.

Lindqvist, O., K. Johansson, M. Aastrup, A. Andersson, L. Bringmark, G. Hovsenius, L.  Hakanson,
       A. Iverfeldt, M. Meili, and B. Timm. 1991. Mercury in the Swedish environment - recent
       research on causes, consequences and corrective methods. Water Air Soil Pollut. 55:1-261.

Majewski, H., S. Brown, R. Evans, H. Freeman, and J. Klaverkamp. 1990.  Responses of kidney,
       liver, muscle, and bone in Atlantic salmon (Salmo solar) to diet and liming in acidic Nova
       Scotia rivers. Can. J. Fish. Aquat. Sci. 47:2441-2450.

Matey, V. 1984.  Comparative analysis of the gill epithelium ultrastructure in perch from basins
       with a different ion composition. Tsitologiya 26:778-782 (In Russian).

McDonald, D. 1983.  The effects of H+ upon the gills of freshwater fish.  Can. J. Zool. 63:691-703.

McMurtry, M., D. Wales, W. Scheider, G. Beggs,  and P. Dimond. 1989. Relationship of mercury
       concentrations in lake  trout (Salvelinus namaycush) and smallmouth bass (Microptems
       dolomieui) to the physical and chemical characteristics of Ontario lakes.  Can. J. Fish.
       Aquat. Sci. 46:426-434.

Meili, M. 1991.   Fluxes, pools, and turnover of mercury in Swedish forest lakes. Water Air Soil
       Pollut. 56:719-727.

Meili, M., A. Iverfeldt, and L. Hakanson. 1991. Mercury in the surface water of Swedish forest
       lakes - concentrations, speciation and controlling factors.  Water Air Soil Pollut.  56:439-
       453.

Mierle, G. and R. Ingram. 1991. The role of humic substances in the mobilization of mercury from
       watersheds.  Water Air Soil Pollut. 56:349-357.

Paulsson, K. and K. Lundbergh. 1991. Treatment of mercury contaminated fish by selenium
       addition. Water Air Soil Pollut. 56:833-841.
                                            107

-------
Rask, M., and P. Tuunainen. 1990.  Acid-induced changes in fish populations of small Finnish
       lakes.  In: Acidification in Finland, P. Kaupii, P. Anttilla, and K. Kenttamies, eds.
       Springer-Verlag, Berlin Heidelberg, pp 911-927.

Reader, J., N. Everall, M. Sayer, and R. Morris. 1989. The effects of eight trace metals in acid soft
       water on survival, mineral uptake and skeletal calcium deposition in yolk-sac fry of brown
       trout, Salmo trutta L.  J. Fish Biol. 35:187-198.

Rosseland, B., O. Skogheim, and I. Sevaldrud. 1986.  Acid deposition and effects in Nordic Europe.
       Damage to fish populations in Scandinavia continue to apace.  Water Air Soil Pollut. 30:65-
       74.

Schindler, D. 1988. Effects of acid rain on freshwater ecosystems.  Science 239:149-157.

Sire, J.-Y. 1986. Ontogenic development of surface ornamentation in the scales of Hemichromis
       bimaculatus (Cichlidae).  J. Fish Biol. 28:713-724.

Spry, D., and J. Wiener. 1991. Metal bioavailability and  toxicity to fish in low-alkalinity lakes: A
       critical review. Environ. Pollut. 71:243-304.

Steingraeber, M., and W. Gingerich. 1991. Hatching, growth, ion accumulation, and skeletal
       ossification of brook trout (Salvelinus fontinalis) alevins in acidic soft waters. Can.  J. Zool.
       69:2266-2276.

Tietge, J., R. Johnson, and H. Bergman. 1988. Morphometric changes in gill secondary lamellae of
       brook trout (Salvelinus fontinalis) after long-term exposure to acid and aluminum.  Can.  J.
       Fish. Aquat. Sci. 45:1643-1648.

Verta, M., J. Mannio, P. livonen, J.-P. Hirvi, O. Jarvinen, and S. Piepponen. 1990.  Trace metal in
       Finnish headwater lakes - effects of acidification and airborne load.  In: Acidification in
       Finland, P. Kauppi, K. Kenttamies, and P. Anttila, eds.   Springer-Verlag, New York, pp
       883-908.

Wiener, J. 1987.  Metal contamination in fish in low-pH lakes and potential implications for
       piscivorous wildlife.  Trans. 52nd NA. Wildl. & Nat. Res. Conf.:645-657.
                                              108

-------
        MORTALITY AND GILL DAMAGE FROM BERYLLIUM IN ACIDIC WATER:
          A COMPARISON OF ACUTE AND CHRONIC RESPONSES IN THREE
                            FRESHWATER FISH SPECIES

                                         by

                       V. E. Matey1, C. H. Jagoe2, T. A. Haines3,
                            V. T. Komov1 and L. Cleveland4
                                    ABSTRACT

       Increased concentrations of toxic trace metals such as Aluminum (Al) are found in
waters acidified by acid deposition, and elevated Beryllium (Be) levels occur in some acidic
waters in eastern Europe.  Beryllium and Al are chemically similar, suggesting that their toxic
effects may be similar as well. We exposed fry of roach and two species of perch to Be in
soft water (calcium [Ca] = 2 mg/L) at two levels of pH (4.5 and 5.5) using static and
flow-through bioassays. In acute toxicity tests with Perca fluviatilis, 10 ug/L or more Be at pH
5.5 produced gill abnormalities including chloride cell hyperplasia and hypertrophy, increased
mucous production, and hyperplasia of the primary lamellar epithelium.  At high
concentrations, we observed fusion and loss of secondary lamellae, progressing to fusion of
adjacent primary lamellae. Less gill damage occurred at pH 4.5, but mortality was much
higher at low Be concentrations.  Roach were killed only when Be was >50 ug/L, regardless
of pH.  Roach gills were damaged by 50 ug/L Be or more at both pH 4.5 and 5.5. With
chronic exposure,  similar abnormalities were caused in P. flavescens gills by 6.25 ug/L or
more Be regardless of pH. The different responses observed may represent interspecies
variation, but were probably influenced by small differences in age among species.
Concentrations of Be similar to those reported in some polluted waters produce gill
pathologies indicative of ionoregulatory stress. The effects of Be and Al are analogous, but
Be is toxic at lower concentrations.
Institute for the Biology of Inland Waters, Russian Academy of Science, Borok, Nekouz,
 Yaroslaval, Russia.
2University of Georgia, Savannah River Ecology Laboratory, PO Drawer E, Aiken SC 29802
 (USA).
3National Biological Survey, Midwest Science Center (formerly  U.S. Fish and Wildlife
 Service, National Fisheries Contaminant Research Center) Field Station,  Department of
 Zoology, University of Maine, Orono ME 04469 (USA).
4National Biological Survey, Midwest Science Center, 4200 New Haven Road, Columbia, MO
 65201 (USA).
                                       109

-------
                                   INTRODUCTION

       The concentration of Be in most surface waters is very low. These levels reflect both
the relative insolubility of BeOH2 at circumneutral pH (Baes and Mesmer 1976) and the
overall scarcity of Be in the earth's crust (Reeves 1986).  Although the average Be content of
soils is low, some soils and soft coals are enriched in Be (Wilber 1980) in portions of eastern
Europe (Vesely et al. 1989).

       Acid deposition increases the mobility of many trace metals (Norton 1982), resulting in
elevated concentrations of toxic trace metals such as Al in surface waters (Cronan and
Schofield 1979). Dickson (1980) showed that cadmium, magnesium, zinc, and lead levels
were elevated in some acid waters,  and these trace elements were apparently mobilized by
acid deposition.  Vesely et al. (1989) reported elevated Be concentrations in acidified waters
in eastern Europe and suggested that Be was mobilized by acid deposition. Both Be and Al
have comparable speciation properties in water (Baes and Mesmer 1976, Vesely et al. 1989).
Given the analogous geochemical properties of Be and Al, Be may also be an important toxic
agent in some acidic waters.

       Few studies have examined toxic effects of Be on fish.  Slonim (1973)  showed that
relatively high concentrations of Be were toxic to guppies Lebistes reticulatis at circumneutral
pH  in  hard water (400 mg/L Ca).  Slonim and Slonim  (1973) found that Be was more toxic to
guppies as Ca levels decreased, but even the lowest  Ca concentrations they tested were
much higher than those present in most acidic waters. In the very dilute waters typical of
acidified lakes and streams, Be would be much more  toxic than previously suspected.  Fish
gills are damaged by Al (Karlsson-Norrgren et al. 1986, Jagoe et al. 1987, Evans et al. 1988,
Tietge et al.  1988), which also causes ionoregulatory  disturbances (Witters 1986, Booth et al.
1988). By analogy,  similar effects might result from Be exposure.
                             MATERIALS AND METHODS

      We performed acute toxicity bioassays at the Institute for the Biology of Inland Waters
in Borok, Russia, in June 1989 to evaluate the toxicity of Be in low pH water and to
determine effects of Be on gill structure (Jagoe et al. 1992). European perch (P. fluviatilis)
and roach (Rutilus rutilus) from the Rybinsk Reservoir were allowed to spawn in artificial
ponds, and fry (perch 0.08-0.15 g and roach 0.7-0.9 g) were collected and acclimated to
reconstituted soft water for 72 hours. This water simulated the ionic composition of
low-alkalinity lakes in the region (Haines et al. 1992), and was prepared by adding salts to
distilled water (final concentrations in mg/L: 2 Ca, 1 sodium, 0.25 potassium, 0.25
magnesium, 1  SO4, and 2.6 chlorine).

      Fish were exposed to Be at two levels of pH in 2 L polyethylene aquaria (n = 10
fish/aquarium), two replicates per treatment, at room temperature  (20-22°C).  Dilute sulfuric
acid was added to the reconstituted soft water to obtain a pH of 4.5 or 5.5, and a BeSO4
solution (nominal concentration 1 mg/mL, measured concentration 0.907 mg/mL, SD = 0.12,
N = 3) was added to yield final nominal concentrations of 0, 10, 25, 50, 100 or  150 ug/L Be.
Controls  (pH 7.0) received neither acid nor Be.  Acute exposures were performed
sequentially, with perch tested first, followed by roach.  Dead fry were counted  and removed
                                        110

-------
at 24-hour intervals, and pH recorded and adjusted if necessary.  The cumulative number of
fish killed by each treatment after 96 hours exposure were compared by analysis of variance.
Separate analyses were conducted for each species and pH level. When significantly
different levels of mortality were found among the Be treatments,  those treatments  which
differed significantly from the control were identified by Dunnetts' t-test (SAS 1988).

       All fish surviving 96 hours were collected and fixed.  During the experiments, some
moribund fry were removed, scored as dead, and fixed to obtain samples for microscopy
from treatments exhibiting  high mortality. Whole fry were fixed in 1% glutaraldehyde and 4%
formaldehyde in 0.1 m HEPES buffer, pH 7.4.  Only fish  that were alive when fixed were
used for microscopic examination.

       To better understand the consequences of chronic exposure to low levels of Be on
fish, we exposed juvenile yellow perch (P. flavescens)  at pH 4.5 and 5.5 for 30 days in May
and June 1990 in laboratory experiments at the Midwest Science Center in Columbia,
Missouri.  Perch  (8-10 months old, average weight 1.5 g) were obtained from the National
Biological Survey, National Fisheries Research Center in La Crosse, Wisconsin.  Exposures
were conducted at 20°C in reconstituted soft water using a flow-through proportional diluter
system as previously described (Cleveland et al.  1986).  Test water contained 2 mg/L Ca,
and was very similar in composition to the water used in  acute toxicity experiments
conducted in Russia in 1989.  The water was acidified to pH 4.5 or 5.5 using a mixture of
sulfuric and nitric acid, and Be was added (as BeCI2) to nominal concentrations of 0, 6.25,
12.5, 25, or 50 ug/L.  Fish  were exposed in 77-L aquaria, and test solutions were renewed at
the rate of about 8 L/hour.   Control treatments  (pH 6.9-7.1)  received neither acid nor Be.
Fish were fed a commercial fish diet ad  libitum three times daily.  In each treatment, pH was
determined daily; oxygen, alkalinity, and conductivity were determined twice each week; and
Be was measured weekly.  Initially, 40 fish were placed in each treatment, and 5 were
removed from each after 5, 15, and 30 days of exposure and fixed for microscopic
examination as described above.

       Fish from both sets of experiments were prepared for scanning electron microscopy
(SEM)  using the same protocols. After at least 24 hours of fixation, the samples were treated
with 1% OsO4 for 1 or 2 hours to increase specimen conductivity,  dehydrated using a graded
ethanol or ethanol-acetone series, then critical-point dried.  European perch and roach fry
were dried whole, mounted on aluminum specimen stubs, and the opercula removed using
fine-tipped forceps to expose the branchial baskets.  For yellow perch, gill arches were
removed and dried individually.  The specimens were spatter-coated with gold or a
gold-palladium mixture, and examined by SEM at accelerating voltages of 5 or 15 kV.
                             RESULTS AND DISCUSSION

ACUTE TOXICITY

       No European perch or roach held at pH 7.0 without Be died during the acute toxicity
experiments.  Exposure to pH 5.5 and Be <50 ug/L did not kill any fish within 96 hours (Figs.
1a and 2a).  Mortality of perch fry exposed for 96 hours to 100 or 150 ug/L Be was
significantly higher than mortality of perch fry exposed to lower concentrations at pH 5.5
                                        111

-------
                                                            a pH 5.5
                                                                     *
                                                                     *
                                                            b.  pH 4.5
                           25             50
                                % MORTALITY
75
100
Fig. 1.  Mortality of European perch fry exposed to beryllium (Be) at pH 4.5 and 5.5. Values
represent the means of two replicates _±_ standard error.  Asterisks indicate significant differences (p
< 0.05) compared to the control (pH 7.2, 0//g/L Be) treatment.
                                                            a. pH 5.5
                                                            b. pH 4.5
                          25             50
                                %  MORTALITY
75
100
Fig. 2.  Mortality of roach fry exposed to beryllium (Be) at pH 4.5 and 5.5.  Values represent the
means of two replicates _+. standard error. Asterisks indicate significant differences (p < 0.05)
compared to the control (pH 7.2, 0 fjg/L Be) treatment.
                                       112

-------
(p < 0.05; Fig. 1a).  Similarly, roach fry were killed by exposure to 100 and 150 |jg/L Be at
pH 5.5 (Fig. 2a) for 96 hours.  Because of variation among the replicates, at pH 5.5 only
mortality at 150 ug/L Be was significantly greater than control mortality in roach (p < 0.05).

       At pH 4.5, mortality of European perch was lowest in  the 0 and 50 ug/L Be treatments
(Fig. 1b), which did not differ significantly from each other or the control treatments (pH 7.0,
no Be). All other Be concentrations tested at pH 4.5 caused significantly higher mortality of
European perch fry after 96  hours (p  < 0.05).  Roach were less sensitive to Be at low pH
than the perch.  Roach fry were killed only at the highest Be concentrations (100 and 150
ug/L) at pH 4.5, and mortality was significantly greater than control mortality (p < 0.05) only
at the highest concentration  tested.

       In European perch, mortality was similar at both levels of pH tested at Be >50 ug/L.
Lower concentrations of Be (<25 ug/L) were more toxic to European perch at pH 4.5 than at
pH 5.5. Apparently the pH 4.5 treatment without Be was also stressful to European perch,  as
over 25% of the fish in this treatment  died. The  reduction in  mortality observed in  50 ug/L
Be, pH 4.5, could have resulted from  a dilution error; however, this situation is unlikely
because reduced toxicity occurred in  both replicates  and the water in these was prepared
independently.  A similar response has been observed in Al exposures; survival and body
ion content of brook trout (Salvelinus  fontinalis) fry at low pH were  reduced at high and  low
Al levels, but were enhanced at moderate levels (Wood et al. 1990).  Polyvalent Al species
present at low pH may stabilize membranes at intermediate concentrations (Baker and
Schofield 1982, Wood et al.  1988a) and Be may function in a similar manner.

        Dissolved Ca clearly influences Al toxicity at low pH  (Brown 1983,  Wood et al.  1990),
probably by modifying gill epithelial permeability.  Previous work on Be toxicity (Slonim  1973,
Slonim and Slonim 1973) showed that guppies in hard water (400 mg/L Ca) at circumneutral
pH had a 96-hour median tolerance limit for Be of 27 mg/L (Slonim 1973).  Toxicity increased
as dissolved Ca concentration decreased; the 96-hour median tolerance limit for guppies was
about 0.2 mg/L Be when Ca = 22 mg/L. In acidified waters, even less Ca is typically present
than used in these previous studies.  Our  studies yielded 96-hour LC50s of  approximately 80
ug/L for European perch fry  and 100  ug/L for roach fry at pH 5.5 when Ca  = 2 mg/L.
GILL DAMAGE DUE TO ACUTE EXPOSURE

       Both juvenile perch and roach held in pH 7.0 water without Be had well-formed gills
with no visible abnormalities (Fig. 3a, b). Secondary lamellae appeared regular in size and
configuration, and individual cells covered with microridges were observed, with little mucous
visible  at higher magnification (Fig.  3c). No gross morphological changes were observed at
pH 5.5 in the absence of Be in gills of European perch (Fig. 4a) or roach (Fig. 4b).  Surface
microridges appeared somewhat less abundant, but no swelling or mucous accumulation was
evident (Fig. 4c).  Exposure to pH 4.5 without Be also had little effect on the gross
morphology of European perch gills (Fig. 4d) or roach gills (Fig. 4e), although we observed
effects typical of low pH stress, such as greater chloride cell numbers. Swelling of mucous
cells indicative of increased mucous production, and many chloride cells with apical crypts
were visible at higher magnification (Fig. 4f).
                                        113

-------
Fig. 3.  Scanning electron micrographs of gills at pH 7.0 with no beryllium: a) European perch, scale
bar = 100/urn; b) roach, scale bar =  1000//m; and c) European perch, scale bar =  100//m.
                                           114

-------
Fig. 4. Scanning electron micrographs of gills of fry exposed to low pH (5.5) without beryllium
(Be): a) European perch, scale bar = 100//m; b) roach, scale bar  = 1000/;m; c) European perch,
scale bar = 100//m; d) European perch, scale bar  = 100//m; e) roach, scale bar =  1000//m; and
f) European perch, scale bar = 10 fjm.
                                          115

-------
       In European perch, exposure to as little as 10 ug/L Be at pH 5.5 caused changes in
gill morphology.  Chloride cell apical crypts were present in perch exposed to 10 ug/L Be
(Fig. 5a), but not in perch exposed to pH 5.5 without Be.  We also found some swelling of
epithelial cells and a reduction in microridges on cell surfaces.  Microridge loss and swelling
of primary lamellar epithelial cells were more marked in European perch after 96 hours
exposure to 25 ug/L Be, pH 5.5.  Exposure to 50 ug/L Be at  pH 5.5 caused hyperplasia  of
the  primary lamellar epithelium in European  perch, which filled in  the spaces between
secondary lamellae, causing the secondary lamellae to appear shorter and less distinct (Fig.
5b).  This hyperplasia  became more severe  at higher Be concentrations; above 100 ug/L Be,
massive epithelial hyperplasia fused adjacent primary lamellae in  perch gills into a continuous
mass (Fig. 5c).  Low Be concentrations produced few effects on roach gills at pH 5.5. At pH
5.5 with 50 ug/L Be, primary and secondary lamellae of roach thickened, the density of
microridges on the surfaces of epithelial cells was markedly reduced, and increased numbers
of enlarged mucous cells appeared on primary lamellae (Fig. 5d).  Gill abnormalities  became
more severe at higher Be concentrations.  Mucous cells were enlarged in gills of roach
exposed to 100 ug/L Be, pH 5.5, and many secretory pores were visible, indicating increased
mucous secretion (Fig. 5e). After 96 hours exposure to 150  ug/L Be at pH 5.5, hyperplasia
of both the primary and secondary lamellar epithelia was pronounced in roach (Fig. 5f),
resulting in gills appearing  greatly thickened and clubbed.  Fusion of adjacent primary
lamellae, as found in European perch fry, was  not observed in roach.

      At pH 4.5,  high mortality occurred in European perch,  but the effects of Be on gill
structure were less striking than at pH 5.5.  European  perch exposed to 50 ug/L Be at pH 4.5
for 96 hours exhibited  swelling of the primary lamellar epithelia, some loss of microridges,
and increased mucous secretion (Fig. 6a).  After exposure to Be at concentrations of 100
ug/L or more at pH 4.5,  areas devoid of microridges were visible  (Fig. 6b) and hyperplasia
caused fusion of adjacent secondary lamellae in a number of regions (Fig. 6c).  In this
species, the hyperplasia at pH 4.5 was much less severe than at  pH 5.5.

       In roach exposed to Be concentrations of 50 ug/L or higher at pH 4.5,  epithelial
hyperplasia occurred.  Primary lamellae became enlarged, with few surface microridges
visible (Fig. 6d).  At 100 |jg/L Be, pH 4.5, hyperplasia was  severe in roach gills; many
secondary lamellae fused as the interlamellar spaces filled in (Fig. 6e).  At 150 ug/L Be, pH
4.5, numerous secondary lamellae were lost in roach gills (Fig. 6f), and epithelial hyperplasia
resulted in thickened, club-like primary lamellae.  Many chloride cell apical crypts and swollen
mucous cells were seen. Although fewer roach than European perch died at  pH 4.5, gill
abnormalities we observed were more severe in the roach at this pH.

       Conversely, Be at pH 4.5 killed most European perch, although they exhibited fewer
gill abnormalities.  Most of the deaths occurred quickly, within 48-72 hours of  exposure.  The
secondary lamellae of European perch dying at nigh Be concentrations were extremely
swollen and irregular,  with  increased amounts of mucous.  This observation suggests Be at
the  low pH damaged epithelial cells and destroyed their function as a water- and ion-tight
barrier.  Disruption of this barrier function likely produced rapid, severe ionoregulatory stress
that killed the fry before compensatory mechanisms could occur,  such  as  epithelial
hyperplasia to increase both chloride cell numbers and the thickness of the diffusion pathway.
                                        116

-------
Fig. 5. Scanning electron micrographs of gills of fry exposed to beryllium (BE) at pH 5.5:
a) European perch, 10//g/L Be, scale bar  = 10//m; b) European perch, 50//g/L Be, scale bar = 10
//m; c) European perch, 100/;g/L Be, scale bar =  100/mi; d) roach, 50//g/L Be, scale bar = 100
   ; e) roach, 100^g/L Be, scale bar = 100/vm; and f) roach, 150//g/L Be, scale bar = 100//m.
                                          117

-------
                                                                                       \
Fig. 6. Scanning electron micrographs of gills of fry exposed to beryllium (Be) at pH 4.5: a)
European perch, 50//g/L Be, scale bar =  10 fjm; b) European perch, 100//g/L Be, scale bar = 10
//m; c) European perch, 150//g/L Be, scale bar = 100jum; d) roach, 50//g/L Be, scale bar =  100
fjm; e) roach, 100 fjg/L Be, scale bar = 100//m; and f) roach,  150jug/L Be, scale bar  = 100//m.
                                           118

-------
EFFECTS OF CHRONIC EXPOSURE

       Survival of yellow perch was not affected by exposure to Be at concentrations up to
50 ug/L at either pH level tested. During the first week of exposure, hyperactivity and
abnormal schooling behavior were observed in yellow perch exposed to 50 ug/L Be at both
pH 4.5 and 5.5, but these effects disappeared as exposure continued.

       The gross morphology of gills was not noticeably different among yellow perch
exposed for 30 days to pH 7.0 (Fig. 7a), pH 5.5 (Fig. 7b), or pH 4.5 (Fig. 7c) without Be.
Both primary and secondary lamellae were distinct and normal in appearance.  Mucous
droplets were more abundant after exposure to the lower levels of pH.  Numerous chloride
cells were present  on primary lamellae at all levels of pH, probably resulting from prolonged
exposure to relatively soft water. Laurent et al. (1985) and Leino et al. (1987) have also
reported chloride cell proliferation in fish held in very soft water.

       In contrast,  30 days exposure to Be produced gill abnormalities, even at the lowest
concentration tested. Compared to controls (pH 7.0, no Be; Fig. 8a), the secondary lamellae
of yellow perch exposed to 6.25 ug/L  Be at pH 5.5 (Fig. 8b) were swollen and wrinkled, and
increased numbers of chloride cells were present in  the secondary lamellar epithelia. The
gills of yellow perch exposed to 6.25 ug/L Be at pH 4.5 were coated with mucous that
obscured much surface detail, and epithelial hyperplasia had reduced the spaces between
adjacent secondary lamellae, making them  shorter and less distinct (Fig. 8c).  This  shortening
and loss of secondary lamellae also occurred at 6.25 ug/L Be  at pH 5.5, and was most
noticeable near the tips of the primary lamellae (Fig. 8d).  At both levels of pH, 6.25 ug/L Be
caused loss of cell  surface microridges (Fig. 8e) and a noticeable increase in chloride cell
surface area. We also observed an unusually high number of mucous cell secretory  pores,
along with bulging,  convex chloride cell  apices in yellow perch  exposed to 6.25 ug/L Be at pH
4.5 (Fig.  8f).

       As in acute  exposure experiments, gill abnormalities in  yellow perch  became more
severe at higher Be concentrations. Thirty  days exposure to 25 ug/L Be at pH 5.5 caused
more  severe  epithelial hyperplasia, causing secondary lamellae to thicken and grow together
(Fig. 9a). This hyperplasia resulted in increased numbers of chloride cells as well (Fig. 9b).
Hyperplasia leading to secondary lamellar loss and primary lamellar thickening also occurred
at 25  ug/L Be at pH 4.5 (Fig. 9c). Abnormalities were not obviously more severe at the
highest Be  level tested (50 ug/L) than at 25 ug/L Be. SecondaryHamellae were also
shortened, thickened, and fused by exposure to 50 ug/L Be for 30 days at pH 5.5 (Fig. 9d).
Exposure to the 50 ug/L Be at pH 4.5 caused similar effects on secondary lamellae (Fig. 9e,
f), and surface microridges were sparse with considerable mucous visible (Fig. 9f).

       The first and fourth gill arches  in young yellow perch always had a row of noticeably
shortened primary  lamellae. In fish exposed to pH 7.0 with no Be, the morphology  of these
short  primary lamellae was very similar to the adjacent longer ones (Fig. 10a).  Exposure to
pH 4.5 without Be for 6 days did not affect the morphology of these shortened lamellae (Fig.
10b),  but they were severely damaged by the presence of Be (Fig. 10c).  To our knowledge,
this report is the first of the presence of these short filaments in the species. Perhaps these
primary lamellae are rapidly growing and developing in young yellow perch, as their size and
structure is similar to that reported in other young fish  (Morgan  1974, EI-Fiky et  al. 1987),
                                        119

-------
Fig. 7. Scanning electron micrographs of gills of yellow perch fry held at various levels of pH
without beryllium (Be); scale bars = 100//m: a) pH 7.0, b) pH 5.5, and c) pH 4.5.
                                            120

-------
Fig. 8. Scanning electron micrographs of gills of yellow perch fry: a) pH 7.0, no beryllium (Be),
scale  bar =  10/vm; b) pH 5.5, 6.25//g/L Be, scale bar = 10/ym; c) pH 4.5, 6.25 fjg/L Be, scale bar
= 10//m; d) pH 5.5, 6.25 fjg/L Be; scale bar = 100/;m; e) pH 5.5, 6.25//g/L Be, scale bar  = 10
fjm; and f) pH 4.5, 6.25/yg/L Be, scale bar = 10^m.
                                           121

-------
 I: V.
Fig. 9. Scanning electron micrographs of gills of yellow perch fry: a) pH 5.5, 25 jjg/L beryllium
(Be), scale bar = 100 fjm; b) pH 5.5, 25 fjg/L Be, scale bar = 10//m; c) pH 4.5, 25 fjg/L Be, scale
bar = 100//m; d) pH 5.5, 50/jg/L Be, scale bar = 100 //m; e) pH 4.5, 50pg/L Be, scale bar =
100/ym; and f) pH  4.5, 50//g/L Be, scale bar =  10 //m.
                                           122

-------
Fig. 10. Scanning electron micrographs of the gills of yellow perch fry showing the shortened row
of primary lamellae on the first and fourth gill arches; scale bars =  100//m: a) pH 7.0, no beryllium
(Be), 30 days exposure; b) pH 4.5, no Be, 6 days exposure; and c)  pH 4.5, 25 fjg/\- Be, 6 days
exposure.
                                            123

-------
although these authors do not mention large differences in size among adjacent rows of
lamellae.  If this situation is true, the rapidly growing structures may be especially susceptible
to pollutant damage, as we observed for Be, and may also be a very sensitive indicator of
pollutant-induced stress.
COMPARISON BETWEEN SPECIES AND pH LEVELS

       The three fish species tested seemed to respond differently to Be. These results may
reflect interspecies differences; for example, roach differs from perch in sensitivity to low pH
(Johansson and Milbrink 1976).  A more thorough comparison of interspecies sensitivity must
consider differences in size and age of fish tested.  European perch fry were somewhat
smaller than roach fry, and both were younger than yellow perch fry. Newly hatched fish and
those just beginning to feed are especially sensitive to low pH and metal-induced stress
(Peterson et al. 1982). As fish grow, they become more resistant.  Thus, the difference in
size and age among the species tested makes evaluation of true interspecies differences in
sensitivity difficult.

       Both Be and Al speciation are pH dependent.  Calculation of equilibrium species
distributions from thermodynamic data indicated that the major species present in both the
acute and chronic toxicity experiments were monomeric Be2+ and BeOH+, depending on pH
(Garrels and Christ 1965, Smith and Martell 1976).  Be2+ accounts for 88% of the total Be
present at pH 4.5, while at pH 5.5, 56% would be in the form BeOhT.  In European perch, the
most striking abnormalities in gills occurred at pH 5.5, when most Be present was in the
hydroxide form. In roach and yellow perch, abnormalities occurred at both levels of pH.
Aluminum-induced gill abnormalities also occur both at pH levels where AIOH species
dominate (Karlsson-Norrgren et al. 1986,  Jagoe et al. 1987) and lower pH levels where AI3+
becomes important (Evans et al. 1988,  Tietge et al. 1988).
COMPARISON OF EFFECTS OF BE AND AL

      Aluminum at low pH causes gill abnormalities, including hyperplasia resulting in
epithelial thfckening, increased mucous production, and changes in chloride cell and mucous
cell number and structure (Evans et al. 1988,  Jagoe 1988, Tietge et al. 1988).  Epithelial
hyperplasia, partly due to increased production of chloride cells, can cause fusion and
disappearance of primary and secondary lamellae: this greatly reduces the functional surface
of the gill (Karlsson-Norrgren et al. 1986, Jagoe et al.  1987).  Gill physiological processes
known to be affected by Al include ion regulation (Witters 1986, Booth et al. 1988) and gas
exchange (Wood et al. 1988b, Playle et al. 1989). Beryllium causes the same gill
abnormalities, which strongly suggests that the same physiological processes are affected by
the two metals.

      The altered chloride cell morphology we observed suggests that Be interferes with ion
regulation, because chloride cells are responsible for ion uptake in freshwater  fish  (Laurent et
al. 1985).  Also, exposure to Be causes apical crypts to develop in chloride cells.  Chloride
cell apical crypts normally occur only in seawater-adapted fish (Foskett and Schelley 1982),
but also occur in freshwater fish experiencing  ionoregulatory disturbances caused by low pH
                                        124

-------
(Leino and McCormick 1984) and At (Jagoe et al. 1987).  Increased chloride cell production
may represent a compensatory response to offset ionic losses if initial exposure to the Be is
not acutely lethal.  However, such hyperplasia involves a trade-off; in exchange for increased
ionic uptake sites, the functional surface area for gas exchange is greatly decreased.

                                   CONCLUSIONS

       Vesely et al. (1989) found Be concentration was highly correlated with decreased pH
in acidified waters in Czechoslovakia. In the waters they sampled, Be >5 ug/L  was found at
27 sites, at concentrations as high as 58 ug/L This study demonstrates that these Be
concentrations would be harmful to fishes, and likely cause gill abnormalities or damage.
Our results also suggest that the effects of Be and Al are analogous.  In some  areas where
Be  is abundant in soils or because of extensive use of lignite coal usage (Wilber 1980), Be
may be an important toxicant whose behavior and effects are still poorly understood.
                                ACKNOWLEDGMENTS

       C.H.J. was partly supported by contract DE-AC09-76SROO-819 between the U.S.
Department of Energy and the University of Georgia's Savannah River Ecology Laboratory.
Financial support was also provided by the U.S. Department of the Interior, Fish and Wildlife
Service; the U.S. Environmental Protection Agency; and by the Institute for the Biology of
Inland Waters, Academy of Sciences of Russia. We thank Boris Flerov, Chief of the
Laboratory of Toxicology and Physiology, Institute for the Biology of Inland Waters, and
Richard Schoettger, Director, National Biological Survey's Midwest Science Center, for
assistance with logistics and facilities.
                                    REFERENCES

Baes, C. F. and R. E. Mesmer.  1976.  The hydrolysis of cations.  John Wiley and Sons
       Publ., New York.

Baker, J. P.  and C. L. Schofield.  1982. Aluminum toxicity to fish in acid waters.  Water, Air,
       Soil Poll. 18:289-309.

Booth, C. E., D. G. McDonald, B. P. Simons and C. M. Wood.  1988. Effects of aluminum
       and low pH on net ion balance in the brook trout (Salvelinus fontinalis).  Can. J. Fish.
       Aquat. Sci. 45:1563-1574.

Brown, D. J. A. 1983.  Effects of calcium and aluminum concentrations on the  survival of
       brown trout (Salmo trutta) at Jow pH.  Bull.  Environ. Contam. Toxicol. 30:582-587.

Cleveland, L., E. E. Little, S. J. Hamilton, D. R. Buckler and J. B. Hunn.  1986.  Interactive
       toxicity of aluminum and acidity to early life stages of brook trout. Trans. Am. Fish.
       Soc.  115:610-620.
                                        125

-------
Cronan, C. S. and C. L. Schofield.  1979.  Aluminum leaching response to acid precipitation:
       effects on  high elevation watersheds in the northeast.  Science 204:304-306.

Dickson, W. T.  1980. Properties of acidified waters. Pages 75-83 in D. Drablos and A.
       Tollan (eds.),  Ecological impact of acid precipitation. Proceedings of an International
       conference, Sandefjord, Norway, SNSF Project, Oslo-As, Norway.

EI-Fiky, N., S. Hinterleitner and W. Wieser.  1987.  Differentiation of swimming muscles and
       development of anaerobic power in the larvae of cyprinid fish (Pisces, Teleostei).
       Zoomorphology 107:126-132.

Evans R. E., S. B. Brown and T. J. Hara.  1988.  The effects of aluminum and acid on the gill
       morphology in rainbow trout, Salmo gairdneri.  Environ. Biol. Fish. 22:299-311.

Foskett, J. K. and C.  Schelley.  1982. The chloride cell: definitive identification as the
       salt-secretory  cell in teleosts. Science 215:164-166.

Garrels, R. M. and C. L. Christ.  1965.  Solutions, minerals and equilibria. Freeman Cooper,
       San Francisco, CA,

Haines, T. A., V. T. Komov and C. H. Jagoe. 1992.  Lake acidity and mercury content offish
       in Darwin National Reserve, Russia.  Environ. Poll. 78:107-112.

Jagoe, C.  H.  1988.  A histological and ultrastructural study of the effects of low pH and
       aluminum upon the gills of Atlantic  salmon.  PhD Thesis, Univ. Maine, Orono.

Jagoe, C.  H., T. A. Haines and D.  R. Buckler.  1987. Abnormal gill development in Atlantic
       salmon (Salmo salai) fry exposed to aluminum at low pH. Pages 375-386 in H.
       Witters and O. Vanderborght (eds.), Ecophysiology of acid stress in aquatic
       organisms, Suppl. 1, vol. 117, Annals Soc. Royal. Zool. Belgium, Antwerp.

Jagoe, C.  H., V. E. Matey, T. A. Haines and V. T. Komov.  1993. Effect of beryllium on fish
       in acid water is analogous to aluminum toxicity.  Aquat. Toxicol. 24:241-256.

Johansson, N. and G. Milbrink. 1976.  Some effects of acidified .water on the early
       development of roach (Rutilus rutilus L.) and perch (Perca fluviatilis L.).  Water Res.
       Bull. 12:39-48.

Karlsson-Norrgren L, I. Bjorklund, O. Ljungberg and P. Runn.  1986.  Acid water and
       aluminum exposure; experimentally induced gill lesions in brown trout, Salmo trutta L.
       J. Fish Dis. 9:11-25.

Laurent, P., H. Hobe and S. Dunel-Erb. 1985. The role of environmental sodium chloride
       relative to  calcium in gill morphology of freshwater fish. Cell Tiss. Res. 240:675-692.

Leino, R. L. and J. H. McCormick.   1984.  Morphological and morphometrical changes in
       chloride cells of the gill of Pimephales promelas after chronic exposure to acid water.
       Cell Tiss. Res. 236:121-128.
                                         126

-------
Leino, R. L, J. H. McCormick and K. M. Jensen.  1987.  Changes in gill histology of fathead
       minnows and yellow perch transferred to soft water and acidified soft water with
       particular reference to chloride cells.  Cell Tiss. Res. 250:389-399.

Morgan, M.  1974.  Development of secondary lamellae of the gills of the trout, Salmo
       gairdneri (Richardson). Cell Tiss. Res. 151:509-523.

Norton, S. A. 1982. The effects of acidification on the chemistry of ground and surface
       waters.  Pages 93-102 in T. A. Haines and R. E. Johnson (eds.), Acid rain/fisheries,
       Am. Fish. Soc., Bethesda, MD.

Peterson, R. H., P.  G. Daye, G. L. Lacroix and E. T. Garside.  1982.  Reproduction in fish
       experiencing acid and metal stress.  Pages 177-196 in T. A. Haines and R. E.
       Johnson (eds.), Acid rain/fisheries, Am. Fish. Soc., Bethesda, MD.

Playle, R. C., G. G.  Goss and C. M. Wood.  1989. Physiological disturbance in rainbow trout
       (Salmo gairdneri) during acid and aluminum exposures in soft water of two calcium
       concentrations.  Can. J. Zool. 67:314-324.

Reeves, A. L.  1986. Beryllium.  Pages 95-116 in L. Friberg, G. Nordberg and V. Vouk
       (eds.), Handbook on the toxicology of metals, vol. II: Specific metals,  Elsevier,
       Amsterdam,  The Netherlands.

SAS. 1988. SAS/STAT Users Guide, Release 6.03 ed.  SAS Institute, Inc., Gary, NC.

Slonim, A. R. 1973. Acute toxicity of beryllium to the common guppy. J. Water  Poll. Cont.
       Fed. 45:2110-2122.

Slonim, C. B. and A. R. Slonim. 1973.  Effect of water hardness on the tolerance of the
       guppy to beryllium sulfate.  Bull.  Environ. Contam. Toxicol.  10:295-301.

Smith, R. E. and A.  E. Martell. 1976. Critical stability constants, vol.4, Inorganic complexes.
       Plenum Press, New York.

Tietge, J. E., R. D. Johnson and H. L. Bergman.   1988. Morphometric changes in gill
       secondary lamellae of brook trout (Salvelinus fontinalis) after long-term exposure to
       acid and aluminum.  Can. J. Fish. Aquat. Sci. 45:1643-1648.

Vesely, J., P. Benes and K. Sevcik.  1989.  Occurrence and speciation of beryllium in
       acidified freshwaters.  Water Res. 23:711-717.

Wilber, C. T.  1980.  Beryllium: a potential environmental contaminant. C. C. Thomas,
       Springfield, IL.

Witters, H. E. 1986. Acute acid exposure of rainbow trout, Salmo gairdneri  Richardson:
       effects of aluminum and calcium  on ion balance and hematology.  Aquat.  Toxicol.
       8:197-210.
                                        127

-------
Wood, C. M., D. G. McDonald, C. E. Booth, B. P. Simons, C. G. Ingersoll and H. L Bergman.
       1988a. Physiological evidence of acclimation to acid/aluminum stress in adult brook
       trout (Salvelinus fontinalis). 1.  Blood composition and net sodium fluxes. Can. J.
       Fish. Aquat. Sci. 45:1587-1596.

Wood, C. M., R. C. Playle, B. P.  Simons, G. G. Goss and D. G. McDonald.  1988b.  Blood
       gases, acid-base status, ions and hematology in adult brook trout (Salvelinus
       fontinalis) under acid/aluminum exposure. Can. J. Fish. Aquat. Sci. 45:1575-1586.

Wood, C. M., D. G. McDonald, C. G. Ingersoll, D. R. Mount, O. E. Johannsson and H. L.
       Bergman.  1990. Whole body ions of brook trout (Salvelinus fontinalis)  alevins:
       responses  of yolk sac and swim up stages to water acidity, calcium and aluminum,
       and recovery effects.  Can. J. Fish. Aquat. Sci. 47:1604-1615.
                                        128

-------
                 IMPACT OF PERMETHRIN ON  ZOOPIANKTON
                 IN HIGH LATITUDE  EXPERIMENTAL PONDS


                                   by


         Thomas W. La Point1,  O.D. Zhavoronkova2 and D.F. Pavlov2



                                ABSTRACT

      This study evaluated the effect of a range  of  concentrations  (0.5-
50 ug liter"1)  of  permethrin on experimental pond zooplankton in northern
USSR.  Each pond was dosed once and the experiment was unreplicated.
Samples were collected from each pond prior to, 7 days and 11 days after
sampling.  High sampling variances precluded measuring statistically
significant differences among the ponds, with the exception of rotifers.
Certain rotifer species appear to be highly sensitive to pyrethroid
insecticides.  The value of this descriptive study is in outlining
future studies of the response of planktonic organisms to one-time or
pulsed dosing of chemicals in experimental ecosystems.  The frequency
and duration of exposure needs to be taken into account relative to the
generation time and reproductive capacity of the organisms under study.
                              INTRODUCTION

      A critical  consideration in measuring  the  effect  of pesticides on
non-target aquatic species is integrating exposure and effect,  relative
to life history characteristics of the exposed organisms.   Trophic
status, including nutrient concentration, suspended particulate,  and
HJ.S.  Fish and Wildlife Service,  National  Fisheries  Contaminant  Research
Center, Columbia, MO (USA) Present address: Department of Environmental
Toxicology and Institute of Wildlife and Environmental Toxicology
(TIWET), Clemson University,  Pendleton,  SC 29670 (USA).
Institute of  Biology of Inland Waters,  Russian  Academy of  Sciences,
Borok, Jaroslavl Region, Russia
                                 129

-------
dissolved and particulate organic matter affects chemical
bioavailability in surface waters (Kaiser 1980, McCarthy & Bartell
1988).   Morpho-edaphic components of the environment (depth, slope,
current, physical-chemical nature of the substrate) also exert a strong
influence on the nature of the biota in aquatic environments (Minshall
1988, Johnson et al. 1993).  Biotic response to a contaminant depends
upon life history characteristics of the individual species (body size,
reproductive potential, and habitat and food requirements)  (Barnthouse
et al.  1987, Barnthouse et al. 1990, La Point 1994).  Further,
consideration must be made for conservatism, as little is known of
inter-specific interactions in planktonic food webs.  Most of the work
to date has dealt with effects on zooplankton  (Liber et al. 1994,
DeNoyelles et al. 1994), benthos (Fairchild et al. 1992, Lozano et al.
1992, Giddings et al. 1994) and fishes (Fairchild et al. 1994,
Barnthouse et al. 1987).  Planktonic and benthic fauna are important in
that they form the base of food webs in natural ecosystems and are
chiefly responsible for decomposition of particulate and in influencing
the overall secondary production within aquatic ecosystems.  How these
not-target groups respond to pesticides will influence rates of
decomposition, primary production,  nutrient cycling, and effects on
organisms higher in the food chain  (Dewey and DeNoyelles 1994).

     Increasingly,  more information is being  collected  on micro  fauna,
particularly the zooplankton.  However,  one group of micro invertebrates
for which there is very little information is that of water mites,
Hydracarina.  Hydracarina are widely distributed all over the world,
except the Antarctic region, and inhabit almost all types of water
bodies  (Gledhill 1985).  The population density of water mites may reach
extremely high levels (Tuzovshiy 1972).   The nymphs and imagoes of water
mites are mainly predators and play a great role in the functioning of
aquatic ecosystems  (Ten Winkel and Davids 1985, Kutikova 1977).  At the
same time, this group of Chelicerata is comparatively poorly studied in
most aspects of aquatic ecology and particularly in aquatic
ecotoxicology.  The number of papers or reports dealing with the
influence of various toxic substances on water mites is comparatively
small and the conclusions made in these papers are often contradictory
(Alexeev 1971, Nair et al. 1977, Stephenson 1982).  For example,  V.A.
Alexeev (1986) stated that Hydracarina are generally resistant to a
variety of toxicants.  On the other hand, it has been shown that they
are very sensitive to toxic action of pesticides and may be useful for
the purposes of bioindication (Nair et al. 1977).

     Synthetic  pyrethroids are  becoming  widely used in  agricultural
regions across the world.  This class of insecticide is characterized by
low water solubility, high lipophilicity, high toxicity to fish and
invertebrates, and relatively low toxicity to mammals.  Because
synthetic pyrethroids are widely used within the USA and Russia,  it is
important that the effect of these chemicals on non-target aquatic
species and ecosystems be studied.   The research described in this paper
was conducted at the experimental ponds of the Institute of Biology of
Inland Waters (IBIW), Russian Academy of Sciences, Borok, Jaroslavl
Region, Russia.  Permethrin  (3-phenoxybenzyl  (IRS)-cis,  trans-3-(2,2-
dichlorovinyl)-2,2-dimethylcyclopropanecarboxylate) was used as the
commercial product Pounce, manufactured by FMC, Inc.  Pounce has 38.47%
active ingredient, 50.7% xylene, and 10.8% related reaction and inert
products.


                                  130

-------
                            MATERIALS AND METHODS

  DESCRIPTION OF STUDY  PONDS

        The  IBIW is located on the  Rybynsk Reservoir in the Jaroslavl
  Region of Russia, roughly 330  km northeast of Moscow at 58  degrees N
  latitude.   The experimental ponds at the IBIW have  a surface area of 800
  m2, volume of 400 m3,  and average depth  of 50 cm.   The ponds are
  characteristic of high-latitude,  oligotrophic ponds  (Table  1).

        Five ponds  were used  in this  experiment:  control, acetone  control,
  and three ponds  treated with permethrin to achieve  concentrations of
  0.5,  5.0 and 50.0 ug  liter"1 (pond numbers 1, 2,  3,  4, and 5,
  respectively).   Acetone was used as the carrier for  permethrin;  1 liter
  was dosed into each pond with  the permethrin by sub-surface application
  (at 15 to 20 cm  depth)  using a hand-held sprayer.   The  concentration of
  permethrin in 1-liter acetone  was 0.52 ml, 5.2 ml,  and  52.0 ml  to
  achieve the 0.5, 5.0  and 50.0  ug liter"1 concentrations, respectively.
  The chemical was mixed into the ponds by transects  taken  across  the
  surface in a boat.  Ponds were treated once, on May  25,  1989.

  RESIDUE ANALYSIS

        Permethrin  samples were intended  to  be analyzed by gas
  chromatography following collection and transport to the  USA.
  Permethrin samples were collected on days 1 and 11  following dosing.
  Collected water  samples were extracted onto C-18 packed glass  columns,
  frozen, and carried to the  National Fisheries Contaminant Research
  Center in Columbia, Missouri,  for analysis.   Unfortunately,  the  samples
  were  not allowed into the country;  hence,  results are in  terms  of
  nominal permethrin concentrations.

Table 1.   Water chemistry of Russian Academy of Sciences Institute (IBIW) ponds at
          Sunuga, Borok,  USSR.
Pond No.
Acetone 0.5 5.0
Control Control pg/1 A"3/l
50.0
Total suspended
particulates, rag/1          1.0          1.9       1.9       5.5        6.0

Total inorganic
phosphorus, pg/1          27.5          6.0       5.0       8.0        8.0

NOj-N, fig/I                0.7          000          0

NOj-N, fig/1               10.0          2.0       2.6       4.0        4.0

Total N, mg/1              0.81         0.70      0.58      0.78       0.73

Biological oxygen
demand,  mg O2/l             1.3          1.3       1.1       1.7        2.5

Total organic
carbon,  mg/1              11.6         11.1      14.0      12.2       11.9

Total suspended
particulate carbon,
mq/1                      0.6          0.3 	nd	1.1	1.2
                                   131

-------
SAMPLING PROCEDURE FOR POND MICROBIOTA

      Pond  microbiota sampled  included  bacteria,  protozoa,  algae,  and
zooplankton.  During the 1-month study, samples were taken 5 times from
each pond  (except the control pond, where samples were taken 4 times).
We sampled once prior to treatment and at 4-day intervals subsequent to
dosing.  Sediments were collected from each pond and subsampled for
chlorophyll a, heterotrophic activity,  and percentage organic matter.

      Bacterial  density was  determined  using  epifluorescence  microscopy
and direct counts on membrane filters stained with erythrosin stain.
Heterotrophic flagellates were quantified using epifluorescence
microscopy with primulin stain.  Rotifera were live-counted in
Sedgewick-Rafter cells.  Zooplankton were sampled using a 10-liter
bucket following standardized techniques for small, shallow water bodies
(Metodika izutchenia 1976).   In each pond, ten 10-liter zooplankton
samples were collected along a transect from the shore to the center for
a total of 100 L water per sample.  Each 100-liter sample was
concentrated using a Juday net with brass screen (No. 73) to volume of
30 ml and preserved in 4 % formalin.  In the laboratory, two subsamples
of either 0.5, 1, or 2 ml (depending on visual density of zooplankters)
were pipetted into Bogorov's counting chamber and zooplankters counted
and identified using a binocular microscope.

      To collect water mites,  the water volume filtered for sampling' was
approximately 4.25 m3per sample.   In all  experimental ponds, except  in
the control, 3 samples were collected during the study period: before
dosing, after 7 days, and after 11 days.  In the control pond, the
samples were collected before dosing and after 11 days.

      Collected water mites  were preserved immediately with Oudemans
fixation liquid.  The identification of water mites' taxonomic status
was conducted in accordance with the classification "of B.A. Wainstein
 (1980).  Water mite species were identified using I.I. Sokolov's  keyword
book  (1940).

DATA ANALYSIS

      For species identification,  we used Manuilova (1964),  Kutikova
 (1970), Bening  (1941) and Opredelitel presnovodnykh bespozvonotcnhykh
 (1977).  For quantifying uncommon  zooplankton species, the entire sample
was enumerated.  All zooplankters were divided into five groups:
Cladocera, Copepoda, Rotatoria, copepod nauplii and copepod copepodites,
and numerical analyses were performed on these groups.

      The pond treatments  were unreplicated.   Comparisons were  drawn
from responses over time for the various taxonomic groups.  Data  were
transformed by logarithms or arc-sine,  as appropriate, to minimize
correlations between the mean and variance of samples.


                                 RESULTS

      Before the permethrin  application,  planktonic communities differed
among the ponds  (Tables 2-4).  At this time, Cladocerans dominated the
zooplankton in all ponds studied; however, density of Cladocera,  as well
                                   132

-------
Table 2.  Density of common cladoceran species (individuals per liter pond water),
SPECIES

Control Pond
Bosmina longirostris
Ceriodaphnia sp.
Diaphanosoma brachiurum
Polyphemus pediculus
Simocephalus vetulus
Total density
Number of species
Acetone Control
Bosmina longirostris
Ceriodaphnia sp.
Polyphemus pediculus
Scapholeberis mucronata
Total density
Number of species
0.5 ug/1
Bosmina longirostris
Ceriodaphnia sp.
Polyphemus pediculus
Simocephalus vetulus
Total density
Number of species
5.0 pg/1
Bosmina longirostris
Ceriodaphnia sp.
Diaphanosoma brachiurum
Polyphemus pediculus
Sida ceystallina
Total density
Number of species
50.0 fig/1
Bosmina longirostris
Ceriodaphnia sp.
Polyphemus pediculus
Total density
Number of species


0

3.6
144.3
0.6
304.8
6.3
465.0
10

172.5
139.6
3.5
4.0
423.5
10

0.3
30.2
40.3
0.3
83.3
11

np
110.4
18.5
6.0
4.8
144.3
9

54.9
307.8
190.2
655.8
10


1

3.6
20.4
3.0
558.0
1.2
600.0
9

40.0
3.7
2.7
0.6
54.3
10

43.5
275.5
8.0
13.5
367.0
9

0.9
0.1
np
np
np
1.0
7

1.2
np
4.2
6.6
4
Density
sample day1
6

12.5
42.5
7.0
6.0
14.5
98.0
9

3.5
3.0
2.6
np
13.2
7

101.5
2.5
36.6
np
142.8
6

112.0
2.5
np
3.5
np
122.5
6

7.0
np
np
7.0
1


10

-
-
-
-
-
-
—

2.6
105.9
0.1
19.7
134.3
9

0.5
40.0
14. t
6.0
66.6
9

42.0
3.0
np
np
1.0
47.3
6

2.0
np
np
2.0
1


15

31.2
np1
19.0
np
np
63.6
6

161.5
26.0
4.0
36.5
238.5
9

3.0
2.0
3.0
9.5
45.0
12

69.0
2.0
np
7.5
0.5
89.5
5

15.6
np
np
15.6
1
'Days  following pond dosing
2np  »  not present.
                                          133

-------
Table 3.  Density (individuals per liter pond water)  of common copepod species
          and of nauplii and copepodites.

SPECIES

Control Pond
Cyclops sp.
Eudiptomus gracilis
Heterocope appendiculata
Mesocyclops leuckarti
Paracyclops fimbriatus
Nauplii
Copepodites
Total density
Number of species
Acetone Control
Acanthocyclops viridis
Eudiaptomus gracilis
E. graciloides
Heterocope apendiculata
Mesocyclops leuckarti
Nauplii
Copepodites
Total density
Number of species
0.5 ug/i
Eudiaptomus gracilis
E. graciloides
Heterocope appendiculata
Mesocyclops leuckarti
Nauplii
Copepodites
Total density
Number of species
5.0 pg/1
Eudiaptomus graciloides
Heterocope
Mesocyclops leuckarti
Nauplii
Copepodites
Total density
Number of species
50.0 uafi
Acanthocyclops viridis
Cyclops strenuus
Mesocyclops leuckarti
Nauplii
t Copepodites
Total density
Number of species
'Days following pond dosing.
2np = not present.
3u = single organism


0

0.6
0.6
np
0.2
np
np
0.1
1.8
8

7.0
3.6
np
4.0
4.0
np
7.4
26.7
5

1.7
np
0.3
np
np
1.2
2.4
3

np
0.3
np
0.6
11.4
20.0
5

3.0
13.2
nd
nd
9.6
16.2
3


4'


1

0.6
np
1.8
1.2
np
np
6.0
3.6
3

0.3
1.2
10.5
6.3
np
0.5
2.4
18.3
4

2.0
1.0
0.2
2.0
4.5
1.1
5.4
7

0.6
1.0
5.5
7.0
5.5
8.2
6

nd
nd
nd
2.4
0.6
0
0


0/1
Density
sample day1
6

np2
0.1
0.1
1.5
np
7.0
15.5
1.7
3

0.1
0.8
1.0
0.3
0.1
4.0
5.0
2.3
5

0.2
np
2.0
np
12.0
18.0
2.2
2

1.0
1.5
3.0
13.0
12.5
5.5
3

nd
nd
nd
1.0
1.0
0
0





10

-
-
-
-
-

-
-
—

np
np
21.0
7.0
0.5
3.0
1.5
28.5-
3

0.2
np
0.6
3.5
14.5
11.5
4.4
4

0.3
0.1
6.3
29.0
18.0
7.6
7

nd
nd
u3
3.0
3.0
0
1





15

9.6
np
np
14.4
7.2
464.4
538.8
32.4
4

np
np
np
np
2.5
121.0
57.0
2.5
1

np
np
np
10.0
8.0
14.0
10.0
1

np
np
36.5
45.0
84.5
40.0
4

nd
nd
9.6
9.6
39.6
9.6
1




-------
Table 4.  Density  (individuals per liter pond water) of common Rotifera species.
     SPECIES
                                                         Density
sample day1
                                                                        10
                            15
Control Pond
  Asplanchna priodonta
  Brachiounus angularis
  B. calcyciflorus
  B. diversicornis
  Filinia longiseta
  Keratella quadrata
  Synchaeta sp.
  Trichocerca cylindrica
    Total density
    Number of species

Acetone Control
  Brachiounus angularis
  B . calcycifl orus
  Euchlanis dilatata
  F. longiseta
  Filinia mayer
  K. cochlearis
  Keratella quadrata
    Total density
    Number of species

0.5
  Asplanchna priodonta
  Euchlanis dilatata
  Filinia longiseta
  F. mayer
  K. cochlearis
  Keratella quadrata
  Lecane luna
    Total density
    Number of species

S.O ug/1
  Asplanchna priodonta
  Brachiounus calcyciflorus
  Conochilus sp.
  Euchlanis dilatata
  Keratella quadrata
    Total density
    Number of species
np
np
np
np
np
np
0.3
np
0.3
1
6.0
np
0.6
np
3.6
4.2
np
np
18.6
8
10.0
35.0
13.0
5.5
5.5
21.0
np
4.5
169.0
19
                           12.0
                          200.4
                            1.2
                          202.8
                          194.4
                            1.2
                             np
                          327.6
                         1348.2
                           18
1.0
1.0
np
np
4.0
2.0
4.0
14.0
7
1.5
2.0
0.5
np
2.0
2.0
7.0
17.5
8
1.0
1.0
np
1.0
1.5
0.5
11.5
20.0
11
np
1.0
1.5
np
2.0
3.0
11.5
23.5
10
32.0
np
20.5
35.0
np
1.0
1.5
117.5
11
0.3
np
np
np
np
0.3
np
1.2
4
1.0
np
4.0
np
np
6.5
np
16.5
7
79.0
np
27.0
190.0
107.5
342.0
3.0
1008.5
18
1.5
8.0
np
5.0
2.5
53.0
11.5
94.0
15
0.5
8.0
1.0
np
np
3.0
5.0
21.0
11
1.5
np
0.2
nd
1.5
2.7
3
2.5
6.5
33.0
1.5
19.0
80.0
15
71.0
66.0
1.5
nd
215.5
603.0
21
44.0
157.0
2.0
64.0
120.0
592.0
21
np
np
np
23.0
18.5
64.5
11
                                         135

-------
Table 4.   (continued)


                                               Density
    SPECIES                                    sample day1
                                01          6          10        15


 50.0
Asplanchna priodonta
Brachiounus calcyciflorus
Euchlania dilatata
F, longiaeta
Filinia nayer
K. cochlearia
Keratella quadrata
Lee ana luna
Total density
Number of species
226.8
np
np
np
np
np
2.7
2.4
236.1
5
139.2
0.6
1.8
np
np
np
0.6
np
144.0
7
30.0
66.5
2.5
18.0
np
11.0
87.5
4.0
283.5
18
8.0
9.0
8.0
6.0
35.0
28.0
50.0
6.0
206.3
18
48.0
32.4
99.6
296.4
np
12.0
9.6
164.5
786.7
17
 'Days following pond dosing.
 2np » not present.
  as  dominant  species  from other orders,  differed among ponds (Figure 1,
  Table  2).  The  lowest initial  density of Cladocera was in pond 3 (83.3
  individuals  per liter),  while  the highest density was in pond 5 (655.8
  individuals  per liter).   Polyphemus pediculus and Ceriodaphnia sp.
  dominated  populations in ponds 1,3,  and 5.   Bosmina longirostris and
  Ceriodaphnia sp.  Were dominant in pond 2,  and Ceriodaphnia sp. and
  Diaphanosoma brachiurum in pond 4 (Table 2).

       Initial copepod density was much  less than  that  of  the Cladocera
  (range 1.8 to 26.7  individuals per liter,  Table 3). Copepod density was
  greatest in  pond 2  and least in pond 1 (Figure 2, Table 3).  Before
  treatment, copepod  nauplii were found only in pond 4 and their density
  was very low, less  than 1 nauplius per liter.  Copepod copepodites were
  present in all  ponds,  but at very low densities  (Table 3).

       Initially,  the highest rotifer  density  was  in pond  5 (Table 4).
  However, this stemmed from one dominant rotifer species,  Asplanchna
  priodonta, which, at 226.8 individuals per liter, was very high compared
  to  the densities of all other rotifers in all ponds.   Typically,  total
  rotifer density was much lower, not exceeding 14.0 individuals per
  liter.  Among the other rotifers, Keratella quadrata was dominant in
  ponds  2-4, while in the control pond, the only species present at this
  time was  Synchaeta  sp.

        Density dynamics  of zooplankton during  the  post-treatment period
  are shown  in Figures 1-4.  Quantitative and qualitative changes occurred
  in all ponds.  By the end of the second week, cladoceran  densities were
  reduced in all  ponds; however, their response dynamics were different
  (Figure 1).   In the control and 0.5 ug/1 permethrin ponds, cladoceran


                                     136

-------
                                       Acetone Control  0.5 //g/L Permethrin
                                         —e—      	•
                                 5.0 //g/L Peimethrin 50.0 /jg/L Permethrin
                         3       6      9      12      15

                             Days Post-dosing
              Figure 1. Number of cladocera per liter in
                       IBIW experimental ponds dosed
                       with permethrin.
density initially  increased,  gradually decreasing to less than pre-
treatment densities.   In  these  ponds,  the initially codominant species,
Ceriodaph/iia sp. and  Polyphemus pediculus,  were replaced by Bosmina
longirostris  (in the  control  pond)  and Simocephalus vetulus (in the 0.5
ug/1 pond).

      In ponds treated with 5.0 and 50.0 ug liter"1  permethrin  (ponds 4
and 5, respectively),  and in  the  control  plus acetone pond (pond 2),
cladoceran densities  decreased  dramatically beginning on day 1.  After
day 3, however, cladoceran density increased in ponds 2 and 4, but not
in 5.  The only cladoceran species that survived in pond 5 was Bosmina
longirostris.  This species became the most common cladoceran at the end
of the observation period in  the  acetone  control and the 5.0 ug liter"1
permethrin ponds  (Table 1).

      Density dynamics of early life-stage copepods  were  generally
similar among the ponds  (Figure 4).   However,  there were marked
numerical differences  among the populations.   In the control pond,
copepod nauplii and copepodites reached densities as high as 464.4 and
538.8 individuals per  liter,  respectively (Figures 3 and 4
respectively).  In all other  ponds,  and especially in the 50.0 ug liter"1
permethrin pond, their densities  were  much  less (Table 3).  However,
after day 15, there was a distinct increase in copepodites and nauplii.
This phenomenon has been  described by  Solomon et al. (1980) and Kaushik
et al.  (1985).  The copepodites and nauplii may show recovery as the
permethrin concentrations in  the  ponds decrease.
                                  137

-------
      By the end of the observation period,  Mesocyclops leuckarti was
the most  common copepod species in all ponds.  However, it was  the  only
copepod species in the acetone control, 0.5 ug liter'1,  and 50.0 ug
liter'1 permethrin ponds,  while in the control and  5.0  ug  liter"1
permethrin ponds,  each copepod population consisted of  four  species.
But, in fact,  the  copepod community was less complex in the  5 ug  liter"1,
because M. Leuckarti  was  accompanied only by single individuals of
Cyclops sp., Acanthocyclops sp.,  and Ectocyclops sp., while  in  the
control pond Cyclops  sp.  and Paracyclops fimbriatus were codominant with
M, leuckarti  (Table 3).

      Rotifers increased in both the control  and  acetone control ponds
(Figure 5); however,  there was an order of magnitude difference in
density between  the two ponds  at  the  end of the experiment (Table 4).  In
the control pond,  four  rotifer species were codominant: Filinia
longiseta, Brachionus angularis,  M.  diversicornis and Trichocerca
cylindrica, while  in  the  acetone  control pond,  F. Longiseta, B.
angularis, and  Euchlanis  dilatata dominated the population.

      Rotifer  density in the 0.5 ug liter'1  permethrin ponds peaked  4
days after permethrin application and  then decreased to pre-treatment
                     Control      Acetone Control  0.5 j/g/L Permethrin

                        5.0 >jg/L Permethrin  50.0 pg/L Permettirin
            0
                                              12
15
                           Days Post-dosing
         Figure 2. Number of copepods per liter  in IBIW
                  experimental ponds dosed with permethrin.
                                 138

-------
                500
                        Control     Acetone Control  0.5 £ig/L Permethrin

                           5.0 //g/L Permethrin 50.0 fjg/L Permethrin
                  0
                                                   12
15
                                 Days Post-dosing

                Figure 3. Number of copepod nauplii per liter
                         in IBIW experimental ponds dosed
                         with permethrin.
levels.  The peaks in density were  primarily a function of the increased
number of Keratella quadrata  in  both ponds,  K. cochlearia and Filinia
mayer in pond 3, and Asplanchna  priodonta and Brachionas calyciflorus in
pond 4 (Figure  4).  By  the  end of the experiment,  K.  quadrata and
Euchlanis dilatata codominated the  rotifer populations in both ponds,
but their densities were  low.  In pond 5,  1 day after permethrin was
applied,  total  rotifer  density was  slightly decreased.  However, the
rotifers gradually increased  and reached  a density of 786.7 individuals
per liter.  The dominant  species at the beginning of the experiment in
pond 5, Asplanchna priodonta,  became much less abundant than Filina
longiseta and Euchlanis dilatata by day 15.   Interestingly, the rotifer
population in pond 5 consisted of 17 species, which was nearly that of
the control pond  (with  18 species)  (Table 4).

      Water mites of Eylais and  Fiona genera  were  dominant  in all  the
experimental ponds.  The  list of water mites collected from the IBIW
ponds is provided in Table  5.  Water mites of other species were found
only in single  exemplars.  Before permethrin was added to any ponds, the
greatest density of water mites  of  Eylais was found in pond 3 and for
Fiona, the greatest density was  in  pond 4.  During the experimental
period, the number of Eylais  increased in ponds I (control) and 2
(control plus acetone)  by 26% and 300%, respectively.

      After permethrin was added, the number  of  water mites of both
genera decreased in all treated  ponds  (except for the Fiona mites in
                                  139

-------
pond 3).  In pond 3,  after  7  days,  the  number of Eylais decreased 32.5%
compared to. pre-treatment levels  and  decreased by 29.1% after 11 days
post-treatment.  In pond 4, density decreased to 20.4% of original at 7
days and thereafter.   The Fiona in  pond 3  increased in number, but this
increase was very slight.   We assume  these mite densities, in fact, did
not change after adding permethrin  at a concentration of 0.5 ug liter"1.
In pond 4, the decline in number  of water  mites of this genus was much
more pronounced.  Seven days  after  treatment,  the number was only 27.5%;
11 days after treatment only  14.5%  compared to pre-treatment period.  In
pond 5, no water mites of either  Eylais and Fiona,  as well as other
genera, were found after pond dosing.


                               DISCUSSION

      Published laboratory and field experiments  have  demonstrated that
synthetic pyrethroids  exert a toxic effect on zooplankton, especially
crustaceans  (Day 1989, Crossland  1982,  Kaushik et al. 1984, Perevoznikov
et al. 1984). Permethrin concentrations as low as 1.0 ug liter"1 and less
were referred as acutely toxic for  the  variety of crustaceans (Day
1989).  However, exposures  simulating natural conditions help to
complete a full evaluation  of the effects  of pyrethroid insecticides on
natural planktonic communities.

      Our  pond experiments have shown that at  a nominal  permethrin
concentration of 0.5  ug liter"1, acutely toxic for a variety of aquatic
invertebrates when tested in  laboratory conditions,  did not produce a
marked direct effect  even on  species  shown to be sensitive to a variety
of toxicants, such as  Ceriodaphnia  (Winner 1988).  Furthermore,  while
              600
                    Control

                       5.0 /jg/L Permethrin 50.0 //g/L Permethrin
Acetone Control  0.5 fig/L Permethrin
                                                 12
                           15
                              Days Post-dosing
              Figure 4.  Number of copepod copepodites per
                        liter in IBIW experimental ponds
                        dosed with permethrin.
                                  140

-------
                              Acetone Control  0.5 ug/L Permethrin
                                —©—
                        5.0/jg/LPermethrin 50.0/jg/LPermethn
                                                12     15
                              Days Post-dosing

              Figure 5.  Number of rotifers per liter in
                        IBIW experimental ponds dosed with
                        permethrin.


5.0 ug liter"1 permethrin led to a  dramatic decrease in density of
cladocerans and copepods,  the  effects were less than would be expected
on the basis of laboratory-measured  responses.

      Our results indicate that, in the control and  acetone-control
ponds, the density  of Eylais water mites  increased during the
observation period.  In  the permethrin-treated ponds at concentrations of
0.5 and 5.0 ug liter"1,  their density decreased (Figure I).  On the other
hand, Fiona density decreased  both in controls and in the 5.0 ug liter"1
permethrin-treated  pond;  however,  in the  treated pond, the effect was
more pronounced.  The concentration  of 0.5 ug liter"1 seems to have no
effect on this genus of mites  (Figure 2).   Eylais mites seem to be more
susceptible to the  permethrin  than the Piona.   However,  at the
permethrin concentration of  50.0 ug  liter"1, all Eylais and Piona water
mites suffered complete mortality.

      Hence,  results of this study demonstrate the susceptibility of
certain water mites  to  permethrin  under  field conditions.  This is in
good accordance with published  results on the toxicity of insecticides
to water mites  (Nair et al.  Stephenson et al.  1986,  Crossland and Wolff
1985).  However, our results are quite different from the results of
investigations on the effects  of industrial pollutants on aquatic mites
(Alexeev 1986).  Evidence  exists to  indicate that the primary reason for
the elimination of  water mites  was the direct toxic effect of
permethrin.  That the elimination  of  this  planktonic invertebrate was
due to an elimination of more  sensitive prey zooplankters seems unlikely
because the mites have  been  shown  to  survive more than 15 to 20 days
when starved  (Zhavoronkova,  unpubl.).
                                  141

-------
   Table 5.  Acari present in Sunuga ponds.
    FAMILY               SPECIES
   Hydrachnidae          Hydrachna ap.

   Eylaidae              Eylaia mullari Koenike, 1897; E.  extender*a O.F.
                        Mull., 1776; E. hamata Koenike, 1897; E. koenikei,
                        1903

   Hydriphantidae         Hydryphantea ap., Hydrodroma ap.

   Limneaiidae           Limnesia ap.

   Hygrobatidae          Fiona coccinea (C.L. Koch), 1836; P. nodata (O.F.
                        Muller), P. variabilia (C.L. Koch), 1836;
                        Hydrochoreutoa ap., Acercua ap.

   Arrhenuridae	Arrhenurus ap.	
      Our experiment was performed  in  mature ponds, all with extensive
macrophyte growth.   The biomass of submerged higher aquatic plants was
high, as was  the  concentration of suspended solids.   The  propensity of
permethrin to adsorb quickly onto solid surfaces  (e.g., suspended
solids, particulate  organic matter, etc.) is well  known  (Muir  et al.,
Sharom and Solomon 1979).

      The IBIW ponds are relatively productive.   High ecosystem
productivity,  with associated high plant surface area and concentration
of particulate organic material,  lessened the bioavailability  of
pentachlorophenol in a mesocosm study  (Robinson-Wilson et al.  1983).
Functional redundancy in the macrophyte community  within  mesocosms also
lessens the potential for  measuring ecological damage (Fairchild et al.
1994).  Empirical laboratory results  (McCarthy and Bartell 1988) show
that  for periphyton,  macrophytes and detritivorous fish,  there is a
decrease in toxic effects  of phenol to daphnia.  The  lessened  toxicity
was directly  related to increased particulate organic matter (POM).
However, for  phytoplankton , zooplankton, benthic  invertebrates and
omnivorous fish,  greater reduction in toxic effects corresponded to
increasing dissolved organic material  (DOM) in the system.  Bacterial
and zooplanktonic growth was dramatically enhanced as a function of
available organic material and, presumably, greater surface area for
microbial activity.

      Higher levels of  primary production corresponded to  lower
pentacholorophenol  (PCP)  residues and toxicity to  fish in a pond-
mesocosm study (Robinson-Wilson et al. 1983).  Yet, lower PCP  toxicity
could not be  ascribed solely to an increase in primary production.
Their results indicated that the larger surface area  available within
                                 142

-------
the macrophyte-dominated ponds contributed to an active epiphytic
community which degraded the PCP.  In a similar manner, the adsorption
of permethrin onto plant surfaces and suspended solids in these ponds
was apparently high and may explain the less pronounced direct toxic
effect of the 0.5 and 5.0 ug liter"1, permethrin concentrations, and the
greater susceptibility of active filter-feeding cladocerans  (Polyphemus
pediculua, Sida crystallina, Diaphanosoma brachlurum) and copepods
(Eudiaptomus gracioloides), while predatory copepods such as Mesocyclops
leuckarti appeared to be much less susceptible to the insecticide.
Among the active filter-feeding zooplankters, such as Polyphemus
pedlculus, Daphnia longispina, Sida crystallina, Eudiaptomus gracilis,
and E. Graciloides, are the most sensitive to the toxic effects of
permethrin.  Of the zooplankton collected in our experiment, Rotifera
were the most resistant to permethrin.  This agrees well with previous
results reported by Kaushik et al.  (1985).  The reasons for such
resistance, as well as factors influencing rotifer population dynamics
as a function of permethrin, were discussed by these authors in detail.

      The  reasons  for the  noted changes  in zooplankton density following
acetone application are not clear.  The decline in Cladocera no doubt
stems from unmeasured pond-to-pond differences.  Such pond-to-pond
differences have recently been documented in a separate study
(Hellenbrant 1994).  Different rates of decline and growth in
zooplankton as a result of permethrin exposure were observed by Kaushik
et al. (1985).  They note zooplankton numerically respond to an
insecticide in two ways, direct reduction from acute toxicity and
indirect numerical responses attributed to release from predatory or
competitive pressures.

      There  is  evidence  that the  recovery  of  zooplankton populations  in
mesocosms is a function of their relatively short generation times, the
number of times they are exposed, and the persistence of the pesticide
in the water column (Fairchild et al. 1991).   In that study, zooplankton
responded much more dramatically to pulsed addition (6 dosing periods
over 12 weeks)  of esfenvalerate than to similar concentratioins dosed
twice over the same period. There was a minimal effect on cladocerans in
ponds dosed one time at 0.1 mg liter"1 carbaryl (Hanazato and Yasuno
1990).  In ponds dosed twice, the cladocerans were able to recover
within to 15 days.  However, in ponds dosed 10 times over a period of 20
days,  each time at the 0.1 mg liter"1 concentration, the cladocerans
failed to recover.

      Recovery  of  copepod  densities  after  permethrin application in this
experiment occurred after 2 weeks,  even in the highest concentration.
The major source of recovery was the appearance of the neonatal
individuals (Figures 3 and 4).( High resistance of eggs of these
copepods to a variety of the environmental stressors has been noted by
previous investigators (Makrushin 1984).  It is important to note  that
another route of recovery for planktonic populations exists, that  of
immigration of smaller or younger forms into the  experimental ponds.
This remains an uncontrolled variable in similar  ponds and mesocosm
studies.

     The effect of pesticides on aquatic  organisms needs to be studied
relative to the potential  for recovery and recolonization of the
species.   An important aspect of disturbance  on communities is the rate


                                  143

-------
at which  the habitat becomes  less suitable  in relation to  the
requirements of the organisms (Southwood  1988).  In effect,  the number of
dosing  intervals and the  concentration of the pesticide become points on
disturbance and adversity axes.   Over sufficient time, disturbance and
periods of adversity have a  strong influence  on  the biotic structure of
communities.  Incorporating  such factors  into future mesocosm studies
will allow a study of species dynamics relative  to the degree and
frequency of disturbances.
                                 REFERENCES

Anderson, R.L. 1989.  Toxicity of  synthetic pyrethroids to freshwater
invertebrates. Environmental  Toxicology and Chemistry 8:403-410.

Associate Committee on Scientific Criteria for Envionmental Quality. 1986.
Pyrethroids: their effects  on aquatic and terrestrial ecosystems. NRCC PubL.
No. 24376. National Research  Council of Canada, Ottawa, Ontario.

Barnthouse, L.W.,  6.W. Suter  II,  A.E. Rosen and J.J. Beauchamp. 1987.
Estimating responses of fish  populations to toxic contaminants. Environmental
Toxicology and Chemistry 6: 811-824.

Barnthouse, L.W.,  O.K. Suter  II and A.E. Rosen.  1990.  Risks of toxic
contaminants to exploited fish populations: Influence of life history, data
uncertainty and exploitation  intensity.  Environmental Toxicology and
Chemistry 9: 297-311.

Bening A. L. 1941. Cladocera  _ (Cladocera of the Caucasus region).
Tbilisi,''is'' publ.house,  383 pp. (In Russian).

Cairns, J., Jr. 1983. Are single  species toxicity tests alone adequate for
estimating environmental hazard?  Hydrobiologia 100: 47-57.

Coats, J.R. and S.P.  Bradbury. 1989. Aquatic toxicology of synthetic
pyrethroid insecticides. Environmental Toxicology and Chemistry 8:359.

Crossland, N.O. 1982. Toxicology  of cypermethrin.   II. Fate and biological
effects in pond experiments.   Aquatic Toxicology 2:205-SSS.

Crossland, N.O. and C.J.M.  Wolff. 1985. Fate and biological effects of
pentachlorophenol  in outdoor  ponds. Environmental Toxicology and chemistry
4:73-86.

Day, K.G. 1989.   Acute, chronic  and sublethal effects of synthetic
pyrethroids in freshwater zooplankton. Environmental Toxicology and Chemistry
8:411-418.

Day, K.E.and N.K.  Kaushik.  1987.  Short-term exposure of zooplankton to the
synthetic pyrethroid, fenvalerate, and its effects on rates of filtration and
assimilation of the alga, Chlamydomonas reinhardii. Archives of Environmental
Contamination and  Toxicology  16:423-432.
                                    144

-------
Day, K.E., N.K. Kaushik and K.R. Solomon. 1987. Impact of fenvalerate on
enclosed freshwater plankton communities. Canadian Journal of Fisheries and
Aquatic Sciences 44:1714-1728.

DeNoyfelles, F., Jr., S.L. Dewey, D.O. Huggins and W.D. Kettle. 1994. Aquatic
mesocosms in ecological effects testing: Detecting direct and indirect effects
of pesticides. Pps. 577-603;  In R.L. Graney, J.H. Kennedy and J.H. Rodgers,
Jr. (Eds). Aquatic Mesocosm Studies in Ecological Risk Assessment. CRC Press,
Inc. Boca Raton, FL.

Dewey, S.L. and F. DeNoyelles, Jr. 1994. On the use of ecosystem  stability
measurements  in ecological  effects testing. Pps.  605-625; In R.L. Graney, J.H.
Kennedy and J.H. Rodgers, Jr.  (Eds). Aquatic Mesocosm Studies in  Ecological
Risk Assessment. CRC Press, Inc. Boca Raton, FL.

Fairchild, J.F., T.W. La Point, J.L. Zajicek,  M.N. Nelson and F.J.  Dwyer.
1991.  Population, community,  and ecosystem-level responses of aquatic
mesocosms to  pulsed doses of  a pyrethroid insectide.  Environmental Toxicology
and Chemistry 11:115-129.

Fairchild, J.F., T.W. La Point and T.R.  Schwartz. 1994. Effects of  an
herbicide and insecticide mexture in aquatic mesocosms.  Archives of
Environmental Contamination and Toxicology 27:527-533.

Ferevaznikov  MA.  , A. A. Salaskin, A. V. Naksuchin, A. P. Lysak. 1988.
(Ecological tolerance of Cladocerans and Copepods to pesticides)  p.p. 42-53
in:  M A. CerevazniSov, A.  N.._Gantverg  (eds)  (The influence of biologically
active chemicals upon hydrobionts). Leningrad,  ''od'' publications  (In
Russian).

Giddings, J.M., R.L. Helm and  F.J. deNoyelles, Jr. 1994. Large-scale outdoor
microcosms: Tools  for ecological assessment of pesticides. Pps. 191-198; In
I.R. Hill, F.  Heimbach, P.  Leeuwangh and P. Matthiessen  (eds.) Freshwater
Field Tests for Hazard Assessment of Chemicals. CRC Press, Inc. Boca Raton,
FL.

Gledhill, T,  1985. Water mites - Predators and parasites. Freshwater Biology
Association Annual Report 53:45-59.

Hamilton, P.B., G.S. Jackson,  N.K. Kaushik, K.R. Solomon, and G.L.  Stephenson.
1988.  The impact of two applications of atrazine on the plankton communities
of in situ enclosures.  Aquatic Toxicology 13: 123-140.

Hanazato, T.  and M. Yasuno.   1990.  Influence of persistence period of an
insecticide on recovery patterns of a zooplankton community in experimental
ponds.  Environmental Pollution 67: 109-122.

Hellenbrandt,  S. 1994. A comparison of zooplankton sampling methods in
evaluating copper sulphate  toxicity in microcosms. Unpublished MS thesis;
University of  Aachen, Germany.
Johnson, R.K., T. Wiederholm and D.M. Rosenberg. 1993. Freshwater
biomonitoring  using individual organisms, populations, and species  assemblages
of benthic macroinvertebrates. In D.M. Rosenberg and V.H. Resh (eds).
Freshwater Biomonitoring and Benthic Macroinvertebrates. Chapman & Hall, NY.


                                   145

-------
Kaiser, K.L.E. 1980. Correlation and prediction of metal toxicity to aquatic
biota. Canadian Journal of Fisheries and Aquatic Sciences 37:211-218.

Kaushik, N.K., G.L. Stephenson, K.R. Solomon and K.E. Day. 1985. Impact of
permethrin on zooplankton communities in limnocorrals. Canadian Journal of
Fisheries and Aquatic Sciences 42:77-85.

Kimball, K.D. and S.A. Levin. 1985.  Limitations of laboratory bioassays: the
need for ecosystem-level testing. BioScience 35(3) 165-171.

Kutikova, L.A. 1970. Kolovratki fauny SSSR _{Rotatorias of the USSR fauna).
Leningrad, ''Nauka"' publ. (In Russian).

Kutikova, K.A. 1977.  (The key book for the determination of the freshwater
invertebrates of the European part of the USSR). L.A. Kutikova' Ya. I. Stov,
eds. Leningrad,''Gidrometxcizdat'' publ. house, (In Russian).

Kutrushin A.V. 1938.  (The use of different Cladoceran species for biotesting)
p. 92-96 in:  Ferevaznikov M A., Gantveog A.N. (eds) The influence of
biologically activites upon hydrobionts). Leningrad,  publications 140 pp.  (In
Russian).

La Point, T.W. 1994. Interpreting the results of agricultural microcosm tests:
Linking laboratory and experimental field results to predictions of effect  in
natural ecosystems. Pages 83-102, In I.R. Hill, F. Heimbach, P. Leeuwangh and
P. Matthiessen (eds.) Freshwater Field Tests for Hazard Assessment of
Chemicals. CRC Press, Inc. Boca Raton, FL.

La Point, T.W., J.F. Fairchild, E.E. Little, and S.E. Finger.  1989.
Laboratory and field techniques in ecotoxicological research: Strengths and
limitations.   In A. Boudou and F. Ribeyre (eds.) Aquatic Ecotoxicology:
Fundamental Concepts and Methodologies, Vol. II.  CRC Press, Boca Raton,
Florida,  pp. 239-255.

Larsen, D.P., F.  deNoyelles, F. Stay and T. Shiroyama. 1986. Comparisons of
single-species, microcosm, and experimental pond responses to atrazine
exposure. Environmental Toxicology and Chemistry 5:179-190.

Liber, K., K.R. Solomon, N.K. Kausbik and J.H. Carey. 1994. Impact of 2,3,4,6-
tetrachlorophenol (DIATOX) on plankton communities in limnocorrals. Pps. 257-
294;  In R.L. Graney, J.H. Kennedy and J.H. Rodgers, Jr.  (Eds). Aquatic
Mesocosm Studies in Ecological Risk Assessment. CRC Press, Inc. Boca Raton,
FL.

Manuilova E.F. 1984.  (Cladocera of the USSR fauna). Leningrad,''Nauka"' publ.
house, 989 pp. (In Russian).

McCarthy, J.F. and Bartell, S.M.  1988. How the trophic status of a community
can alter the bioavailability and toxic effects of contaminants.  In J.
Cairns, Jr.  and J.R. Pratt (eds.) Functional Testing of Aquatic Biota for
Estimating Hazards of Chemicals.  ASTM STP 988. American Society of Testing
and Materials, Philadelphia, PA.  pp. 3-16.

Methods of the investigation of the inland water biogeohydrobionts). 1976.  F.
O.i-Boltovskyup,  ed. Leningrad,  ''Nauka1" publ. house,  240 pp.(In Russian).

                                    146

-------
Minshall, G.W. 1984. Aquatic insect-substratum relationships. Chapter 12 In
V.H. Resh and D.M. Rosenberg (eds). The Ecology of Aquatic Insects. Praeger,
NY.

Muir, D.G.C., G.P Rawn and N.P. Grift. 1985. Fate of the pyrethroid
insecticide deltamethrin in small ponds: a mass balance study. Journal of
Agricultural and Food Chemistry 33:581-591.

Robinson-Wilson, E.F., T.P. Boyle, and J.D. Petty.  1983.  Effects of
increasing levels of primary production on pentachlorophenol redisues in
experimental pond ecosystems.  In W.E. Bishop, R.D. Cardwell, and B.B.
Heidolph  (eds.) Aquatic Toxicology and Hazard Assessment: Sixth Symposium.
ASTM STP 802.  American Society of Testing and Materials, Philadelphia, PA.
pp. 23;9-251.

Solomon, K.R., K. Smith, G. Guest, J.Y. Yoo and N.K. Kaushik. 1980. Use of
limnocorrals in studying the effects of pesticides in the aquatic ecosystem.
Canadian Technical Reports on Fisheries and Aquatic Sciences 975:1-9.

Southwood, T.R.E.  1988.  Tactics, strategies and templets.  Oikos 52: 3-18.

Stephenson, G.L., N.K. Kaushik, K.R. Solomon and K. Day. 1986. Impact of
methoxychlor on freshwater communities of plankton in limnocorrals.
Environmental Toxicology and Chemistry 5:587-603.

Ten Winkel, E.H. and C. Davids. 1985.rBioturbation by cyprinid fish affecting
the food availability for predatory water mites. Oecologia 67:218-219.

Wainstein, B.A. 1980. (Key to larval water mites) Institut Biologii
Vnutrennikh Vod. (Academiia Nauk SSSR) 238 pp. (In Russian)

Winner R.W. 1938. Evaluation of the relative sensitivity of 7-d Daphnia magna
and Ceriodaphnia dubia toxicity tests for cadmium and cadmium
pentachlorophenate. Environmental Toxicology and Chemistry 7:153-159.

Yasuno, M., T. Hanazato, T. Iwakuma, K. Takamura, R. Ueno and N. Takamura.
1988. Effects of permethrin on phytoplankton and zooplankton in an enclosure
ecosystem in a pond. Hydrobiologia 159:247-258.
                                   147

-------
      THE EFFECT OF HEAVY METALS AND CHLORPYRIFOS. SEPARATELY AND
     IN COMBINATION. ON A CONTINUOUS FLOW MESOCOSM AQUATIC SYSTEM

                                         by

         G. A. Vinogradov1, F. Stay2, P. P. Umorin1, A. S. Mavrin1, A. K. Klerman1,
         E. I.  Koreneva1, S. A. Kurbatova1, I. O. Solntseva1 and G.  I. Vinogradova1
                                    ABSTRACT

       Pollutants selected for testing were heavy metals and the pesticide chlorpyrifos. A
flow-through aquatic system was found to be useful to test biological effects of these
toxicants separately and in combination.  The continuous flow is more natural, as pollutants
are usually being continuously added to a waterbody. This ecosystem-level testing is seen
as the only strategy available to obtain ecologically relevant information.  A good fit of
predicted and observed effects shows that mathematical modeling can be applied to  such
systems.
                                  INTRODUCTION

      Test systems at different levels of biological order can be used to identify potential
impacts of pollutants in an attempt to predict possible damage that would occur in natural
ecosystems. Ecosystem-level testing is advocated as the only strategy available to obtain
ecologically relevant information, and a variety of model ecosystems are used in studies of
aquatic pollution (Andrushkaitis et al.  1989, Lundgren 1985).

      This research was carried out by the Experimental Ecology Laboratory of the Institute
for the Biology of Inland Waters,  Russian Academy of Science, as a part of the joint USA and
USSR project on the evaluation of techniques used to predict effects of environmental
pollutants.  In 1986, the Environmental Research Laboratory at Duluth,  Minnesota, performed
studies on effects of chlorpyrifos on a natural aquatic system using littoral enclosures.  We
developed a flow-through mesocosm test system using 1.5 m3 tanks designed to measure
toxicant effects on main ecosystem structure variables.  The rationale behind this test system
is that it is large enough to include larger organisms, such as fish, and it is more easily
Institute for the Biology of Inland Waters, Russian Academy of Science, Borok, Nekouz,
 Yaroslaval, Russia.
2U.S. Environmental Protection Agency, Environmental Research Laboratory, Duluth, MN
 (USA).
                                        148

-------
controlled and monitored than bigger systems such as ponds. The continuous flow regime
was chosen because flow-through systems tend to approach a steady state depending on
system inputs and, therefore, are easier to control and to model mathematically.  The other
factor which contributed to the decision to work with this system was to compare different
approaches to the same problem used in the USA and  USSR.  Since many contaminants
contribute to the pollution of natural aquatic systems, it was of interest to look at effects of
the combination of some pollutants.  Heavy metals and the pesticide chlorpyrifos were
selected  because of the high priority status given these substances in both countries.
                                      METHODS

       Twenty-four fiberglass tanks half filled with water were placed into a pond and served
as a thermostat. The experiment was conducted at natural illumination and temperature.
The tanks were filled to a volume of 1.5 m3 with pond water containing natural plankton
assemblages, and a 5 cm thick layer of bottom sediments taken from the same pond was
added to each tank.  The pond water first passed through plankton gauze net (mesh size 80
/vrn) and was then continuously pumped into the  tanks at a dilution rate of 0.33/day. The
same kind of net was installed at the outlet of each tank to prevent washout of organisms
bigger than 80 //m.

       The experimental design consisted of eight treatment tanks, each with three
replicates.  Treatment tanks (the first four tanks were without fish) were as follows: 1 =
chlorpyrifos (Ch), 2 = heavy metals (HM),  3 = chlorpyrifos and heavy metals (Ch + HM),  4 =
control (C); the second four, tanks 5-8, received the same treatments but with fish (+ F). The
HM included zinc, cadmium, copper, and mercury in concentrations of 150.0,  15.0, 7.5, and
1.5 /yg/L, respectively. The mixture of the HM was continuously supplied to the tanks by a
diluter designed by G. A. Vinogradov (Vinogradov and Tagunov 1989) to provide the above
concentrations in the total inflow.  These concentrations were selected taking into account
our mesocosm experiments of 1989 which revealed different crustacean species' sensitivity
to the heavy metals used in this research. Chlorpyrifos was added manually once a week to
attain a concentration in the tank of 0.1 //g/L. This concentration was chosen on the basis of
the results of preliminary 10-day experiments with zooplankton  communities, which showed
that this chemical causes statistically significant changes in the  abundances of crustaceans
without elimination of any of the zooplankton groups.  Fish treatments initially contained 10
bream (Abramis brama) underyearlings  (each weighing 0.4 g on the average) per tank.
According to preliminary  10-day experiments, such a density offish exerted the same
influence on zooplankton as the toxicants.

       To measure effects of the toxicants, we monitored weekly what we believe to be
principal structural ecosystem variables. They included phytoplankton (chlorophyll a
concentration), bacterioplankton (number of cells), and zooplankton (number and size
structure of crustaceans and rotifers). Concentrations of principal ions, nutrients, dissolved
oxygen, and organic matter were determined to provide information about general conditions
under which the experiments were performed. No significant differences were found in these
parameters among treatments and over the time period. Average values are given in Tables
1-4.  Fish studies included determinations of linear and  mass growth of fish and their
resistance to flow.
                                        149

-------
TABLE 1.  CONTENTS OF PRINCIPAL IONS IN THE MESOCOSM WATER, MMOL/L
 CaH
           Na+
                   HCO
                    SO
                    Cl
                     PH
 0.43
1.04
0.21
0.10
1.70
0.22
 0.15
6.9-9.0
TABLE 2.  CONCENTRATIONS OF NUTRIENTS (MG/L), DISSOLVED ORGANIC MATTER (DOM,
MG/L), AND DISSOLVED OXYGEN (DO, MG/L) IN THE MESOCOSM WATER
 Nitrate     Nitrite
          Ammonium     P miner   P total
                              DOM
                             DO
                             max1
                              DO
   50
 1.0
  30
    20
   40
  10
9.0-12.8   6.8-9.0
 fDO max measured at 1 pm; DO min measured at 5 am.
TABLE 3.  TEMPERATURE REGIME IN MESOCOSMS

T min
T max
25/07
16.7
20.6
02/08
15.8
19.3
09/08
18.3
23.7
16/08
17.2
20.8
23/08
18.0
17.3
30/08
10.1
13.2
12/09
9.6
11.3
TABLE 4.  AVERAGE MEASURED CONCENTRATIONS OF HEAVY METALS IN THE CONTROL AND
HEAVY METAL TREATMENTS, MG/L (MEANS AND STANDARD ERRORS)
Control
                      Heavy metal treatments
 Zn
         Zn
              Cd
                 Cu
                   Hg
14.9  1.9
       103.4 17.2
                                       < 1
 'Below analytical detection level.
                                     150

-------
       To analyze experimental results, we used mostly multivariate statistics (Dillon and
Goldstein 1984).  Differences among treatments and effects were assessed using 2-way
ANOVA with repeated measures (Vanni 1986). To predict major changes in the principal
ecosystem living components in response to the influence of some anthropogenous pollutants
(such as heavy metals and pesticides), a mathematical model was developed.
                                MODEL DESCRIPTION

       Because of the lack of space, only some principal aspects of the model are described
here. The model includes the usual components of trophic levels found in any ecosystem:
higher aquatic plants, planktonic and periphytic (or benthic) algae and bacteria, herbivorous
and carnivorous zooplankton, and fish.  Dead organic matter is represented by detritus
(particulate matter), dead plants, and fish.

       For the purpose of ecosystem simulation, various biotic components  have been
subdivided into groups, based upon size, trophic function, and sensitivity to a toxicant.  For
instance, copepods are divided into two trophic groups, herbivorous and carnivorous.
However,  to assess their sensitivity to the toxicant, they are classified as one group.  Further,
algae are  divided into two  size classes to measure growth rates and consumption by
herbivores; however, they  are combined into one group to assess their sensitivities to the
toxicant. Herbivorous copepods and cladocera are lumped into  one group of "herbivorous
Crustacea,"  but they are split into  two groups according  to their sensitivity to toxicants (i.e.,
cladocera  are generally much more sensitive to toxicants than are copepods). This
generalization  is carried  through into the model equations (Baudouin and Scoppa 1974,
Anderson  1982, Kaushik et al. 1985).

       Mathematically, the model  is essentially a system of differential equations. These are
mass balance equations describing the flow of one chemical element through various trophic
levels in aquatic systems.  It is nitrogen in the model, but could be applied to any other
biostructural element without changing the model structure. The model is based upon the
following assumptions: growth rates of the organisms are functions  of food substrate
concentrations which follow generally accepted Michaelis-Menten kinetics.  Equations for the
living components are Verhulst-Pearl equations, whereby the natural mortality of the
organisms is proportional to the  square  of their biomass.  A limiting substrate for the
planktonic and periphytic algae and bacteria is assumed to be mineral nitrogen dissolved in
the water.  Higher plants are assumed to take their nitrogen from a source external to (not
part of) the system (e.g., deeper layers  of the bottom sediments).  However, the material is
added to the system during lysis of dead bodies; therefore, this assumption is valid for only
rooted submersed vegetation.

       As  for benthic and periphytic organisms, their carrying capacities are assumed to be
functions of the area available for  their settling, i.e., surfaces of bottom sediments, higher
plants, and side walls (or banks) of the containers.

       Generally, the equations  describe biomass growth, food consumption (including
grazing and  predation), natural mortality, and the effect of chemical toxicity to these biotic
                                        151

-------
components.  Coefficients of feeding selectivity are incorporated using the entire diet, with
proportions established based upon percentages in the diet.

       At this initial stage, the model is meant to describe changes over relatively short time
periods, usually used in experimental studies (not more than 2 months).  For these shorter
time periods, seasonal periodicity in the organisms' reproduction, and changes in the light
and temperature regime can be neglected.  The model will serve to describe batch (static)
and flow-through systems.

       Model parameters used in simulations were taken from the literature (Eppley et al.
1969, Buhr and Miller 1983, Kokova 1982, Kruychkova 1974, Zaika 1972, 1983).  It should be
noted,  however, that cited literature sources do not usually contain the needed parameters;
the latter had to be recovered  from the quantitative data provided in the sources.  With these
parameters, the model predicted correctly (qualitatively and in many cases, quantitatively) the
effects of factors applied in the experiments.  After experimental data were obtained, the
model was calibrated to achieve a better fit of predicted and observed effects. Tables below
contain model simulation data  after such calibration.
                             RESULTS AND DISCUSSION

       Here we present the results of observations of the main ecosystem variables over a
period  of 1  month (from July 27 until August 27) during which time toxicants were applied to
the experimental mesocosms.  The  period of recovery that followed is not considered. The
most pronounced changes due to toxicant application occurred in the zooplankton
community.

       The natural seasonal course of zooplankton, as observed in the control mesocosms,
was characterized by an increase in Cladocera from 25 to 55% of total zooplankton
abundance. At the same time, Copepoda decreased  from 72 to 38%.  The abundance of
Rotifera only slightly increased during the observation period (from 3 to approximately 15%).

       Effects of the toxicants, separately or in combination, are shown in Tables 5-9. They
were calculated as differences between the  means of treatments and corresponding controls,
so that positive  values mean an increase and negative values a decrease compared to the
control. Fish are also considered as a factor which influences other ecosystem variables.

       Since biomasses of the mesocosm living  components were not estimated and the
model provided data in terms of biomass, we do not compare here the time  courses of the
components in the model and  the experiments.  Still, we can assess effects  of factors within
the model and the experiments and compare these effects.  Comparisons are valid provided
there are no significant differences in the size structure of the compared populations among
treatments, so that linear relationships between biomass and number may be assumed.  This
was not the case with crustaceans; therefore, model values of their biomasses were
converted into organism numbers for assessments of the effects, taking into account the size
structure of their populations.
                                        152

-------
TABLE 5. EFFECTS (PERCENT OF CONTROL) OF INDIVIDUAL FACTORS UPON MAIN ECOSYSTEM
VARIABLES. UPPER LINE FOR EACH CELL REPRESENTS EXPERIMENTAL DATA, LOWER LINE =
COMPUTER SIMULATION, SIGN. 1  =  SIGNIFICANCE LEVEL
Ecosystem
variables
Cladocera
abundance
Copepoda
abundance
Rotifera
abundance
Algae
abundance
Bacteria
abundance
Number of
Cladocera
Copepoda
Rotifera
Heavy metals
Effect
-44.76
-36.30
16.78
11.39
-42.66
-29.22
-4.25
3.94
9.24
8.33
species
-5.64
-1.87
-11.70
Sign. 1
<0.01
0.07
<0.01
0.35
0.25

0.20
0.35
<0.05
Chlorpyrifos
Effect
-53.24
-57.54
33.32
21.65
37.31
11.34
24.29
4.43
12.60
13.41

-26.15
-18.44
-3.02
Sign. 1
<0.01
<0.01
<0.01
<0.05
0.14

<0.01
<0.01
0.31
Fish
Effect
-44.26
-50.25
53.34
49.29
75.36
73.16
60.04
13.57
43.80
27.26

-6.64
-5.88
24.46
Sign. 1
<0.01
<0.01
<0.01
<0.01
<0.01

<0.05
0.08
<0.01
       Analysis of the whole body of data (Tables 5-7) has shown that the direct individual
effect of heavy metals and Chlorpyrifos is a reduction in abundance of the most sensitive
species due to acute toxicity.  This was indicated by an overall decline  in abundance of
cladocerans.  These crustaceans have been found to be the species most sensitive to
toxicants in many other studies (Baudouin and Scoppa 1974, Hurlbert 1975, Kaushik et al.
1985). In our experiments, only three species persisted in the toxicant treatments.  These
were Ceriodaphnia quadrangula, Bosmona longirostris, and Daphnia longispina.

       Judging by a reduction in the number of species, both types of pollutants are directly
toxic to copepods but to a lesser extent than to cladocerans. On the whole, Chlorpyrifos
appeared to be more toxic than the heavy metals as seen by the greater reduction in
cladoceran  abundance and a greater and highly statistically significant reduction in species
richness of  both groups of crustaceans.  The mathematical model used here showed a
decrease in the biomass of copepods exposed to both toxicants.  In real experiments,
however, abundance (number of individuals) of the copepods significantly increased. An
analysis of the size structure of the copepod population revealed  that elimination  of larger
species and specimens occurred, and the ratio  of large to small individuals suffered an
almost 5-fold decrease. One of the events was complete  elimination of the adults of
Mesocyclops crassus and M.  oithonoides.  The copepod population consisted mostly of
nauplii and  copepodites rather than adults, probably because the toxicants induced some
                                       153

-------
hindrances to growth or moulting.  It is only after model biomass data were converted into
organism numbers that a good fit of the observed and  predicted effects of factors was
obtained.

       As seen from Table 5, the mixture of heavy metals in the concentrations used is
probably directly toxic to  rotifers because the effect is great and statistically significant.  In the
literature (Moore and Ramamoorthy 1984), rotifers have been noted to be among the
organisms most sensitive to heavy metals. The toxic effect of chlorpyrifos is insignificant
judging by a decrease in species richness and by the fact that their abundance did not
diminish.

       Both types of toxicants apparently were  not directly toxic to algae and bacteria, as
negative effects are small and statistically insignificant.  Concentrations of toxicants used  in
the experiments were below acute toxicity and growth inhibition concentrations reported in
the literature for algae and bacteria (Hurlbert et al. 1972,  Roberts and Miller 1970, Maly and
Ruber 1983, Stratton et al. 1980, Moore  and Ramamoorthy 1984).
TABLE 6. EFFECTS (PERCENT OF CONTROL) OF A COMBINATION OF FACTORS UPON MAIN
ECOSYSTEM VARIABLES. UPPER LINE FOR EACH CELL REPRESENTS EXPERIMENTAL DATA,
LOWER LINE = COMPUTER SIMULATION, SIGN. 1 = SIGNIFICANCE LEVEL, INTER. =
INTERACTION
Ecosystem
variables
Cladocera
abundance
Copepoda
abundance
Rotifera
abundance
Algae
abundance
Bacteria
abundance
Number of
Cladocera
Copepoda
Rotifera
Heavy metals
Effect
-72.05
-75.21
54.21
35.52
-22.35
-21.34
18.20
8.57
-4.91
23.77
species
-30.36
-19.49
-14.33
Sign. 1
<0.01
<0.01
<0.01
0.11
0.36

<0.01
<0.01
<0.05
+ chlorpyrifos
Inter.
25.85
18.63
4.12
2.47
16.99
-3.46
-2.14
0.89
-1.50
1.33

1.43
0.82
0.39
Sign. 1
0
>0
0
0
>0

0
0
0
.27
.05
.07
.26
.50

.25
.30
.46
Heavy metals + fish
Effect
-61.99
-73.09
86.36
67.41
0.68
12.12
63.22
18.53
56.81
39.72

-10.78
-7.63
10.58
Sign. 1
<0.01
<0.01
>0.50
<0.01
<0.05

0.09
0.16
0.15
Inter.
22.93
13.46
16.24
6.72
-32.03
-31.81
7.43
1.04
3.71
4.11

0.50
0.13
-2.17
Sign. 1
0.29
0.35
0.47
<0.01
>0.50

0.46
>0.50
0.35
                                        154

-------
TABLE 7.  EFFECTS (PERCENT OF CONTROL) OF A COMBINATION OF FACTORS UPON MAIN
ECOSYSTEM VARIABLES.  UPPER LINE FOR EACH CELL REPRESENTS EXPERIMENTAL DATA,
LOWER LINE = COMPUTER SIMULATION, SIGN. 1 = SIGNIFICANCE LEVEL, INTER. =
INTERACTION
Ecosystem
variables
Cladocera
abundance
Copepoda
abundance
Rotifer a
abundance
Algae
abundance
Bacteria
abundance
Number of
Cladocera
Copepoda
Rotifer a
Chlorpyrifos + fish
Effect
-72.63
-81.52
103.92
81.13
91.19
85.57
94.32
19.46
24.98
48.49
species
-30.63
-22.66
19.46
Sign. 1
<0.01
<0.01
<0.01
<0.01
0.28

<0.01
<0.01
<0.05
Inter.
24.86
26.28
17.26
10.18
-21.48
-31.81
9.69
1.45
-23.40
7.79

1.16
1.67
-1.97
Sign. 1
0.27
0.21
<0.01
<0.01
>0.50

0.25
0.10
0.28
Heavy metals + Chlorpyrifos + fish
Effect
-83.75
-88.20
136.72
123.43
-11.38
35.18
78.72
25.74
17.05
70.69

-39.11
-21.13
-1.94
Sign. 1
<0.01
<0.01
0.42
<0.01
0.23

<0.01
<0.01
0.41
Inter.
58.41
55.89
33.28
41.07
-81.39
-20.09
-1.66
3.79
-23.40
21.67

-1.68
5.08
-11.67
Sign. 1
0.11
0.16
0.06
<0.01
>0.50

0.16
<0.05
0.48
       When fish are considered as another factor (actually predation by fish), the effect is,
superficially, similar to effects of the toxicants; although some principal differences emerge.
Fish presence caused almost the same magnitude of decline in cladoceran abundance and in
the biomass of copepods as did toxicants, but their effect on crustacean population size
structure was more dramatic.  Fish predation appeared to be much more size selective than
toxicity: the ratio of small cladoceran forms to larger ones in the toxicant treatments
increased only slightly over the control ratio,  while in the fish treatment,  the large forms were
near complete extinction.  As for copepods, this ratio in the presence of fish was almost twice
as great as in toxicant treatments.  On the whole,  the main effect of fish is selective
elimination of large crustacean forms.

       Concerning direct effects of fish on rotifers, algae, or bacteria, it  seems from the
information stated, that fish did not feed on any of them, as their diet consisted  mostly of
crustaceans.

       Several indirect effects of factors applied here were observed. Indirect or secondary
effects may be described  as changes in ecosystem variables  resulting from changes caused
by direct toxicity to some most sensitive populations (or by predation  in  case of fish). The
                                        155

-------
decline in the cladoceran population due to toxicant effects caused a significant increase in
the number of algae (mostly small forms as the model showed), which comprise the main
food item for herbivorous copepods and for immature stages of the predators (Wylie and
Currie 1991).  This led to release of these forms of copepods (and rotifers in the case of
chlorpyrifos) from competition pressure and to an increase in their abundance.  The model
and experimental observations showed an increase in small herbivorous copepods; predators
changed very little. Correlation analysis of the whole body of experimental data revealed a
negative relationship between abundances of herbivorous copepods and cladocerans
(r = -0.82, p < 0.05), which supports the paradigm of their competition.

      Secondary effects offish addition also included increases in abundances of algae,
bacteria, rotifers, and copepods.  These effects were even more pronounced than those of
the toxicants because in addition to cladocerans, fish also greatly reduced large copepods
which are mostly predatory.  Therefore,  rotifers and herbivorous copepods were released not
only from competition but also from predation.  Both invertebrate predation and competition
have been found to be very important in determining freshwater zooplankton community
structure (Vanni 1988, Soto and Hurlbert 1991) and contribute to the scarcity of small
TABLE 8. EFFECTS (PERCENT OF CONTROL) OF INDIVIDUAL FACTORS AND THEIR
COMBINATION UPON MAIN ECOSYSTEM VARIABLES WITHOUT FISH. UPPER LINE FOR EACH
CELL REPRESENTS EXPERIMENTAL DATA, LOWER LINE = COMPUTER SIMULATION, SIGN. 1 =
SIGNIFICANCE LEVEL, INTER. = INTERACTION
Ecosystem
variables
Cladocera
abundance
Copepoda
abundance
Rotifera
abundance
Algae
abundance
Bacteria
abundance
Number of
Cladocera
Copepoda
Rotifera
Heavy
Effect
-51.78
-35.38
36.78
13.15
-8.20
-25.34
32.44
7.34
7.99
18.44
species
-8.82
-2.41
1.46
metals
Sign.
1
<0.01
0.50
Chlorpyrifos
Effect
-55.82
-58.02
35.18
28.71
46.29
38.14
13.14
11.78
-17.63
33.33

-24.32
-23.02
-14.77
Sign.
1
<0.01
<0.05
<0.01
0.26
0.19

<0.01
<0.01
0.09
Heavy metals
Effect
-75.81
-75.92
76.55
46.71
15.24
3.70
43.77
20.15
-10.90
60.34

-30.17
-24.65
-12.90
Sign.
1
<0.01
<0.01
<0.05
0.10
0.32

<0.01
<0.01
0.20
+ chlorpyrifos
Inter.
31.79
17.48
11.59
4.85
19.25
-9.44
-1.80
1.02
-1.26
8.57

2.98
0.78
0.40
Sign.
1
0.28
>0.50
0.17
0.34
>0.50

0.16
0.42
>0.50
                                        156

-------
TABLE 9. EFFECTS (PERCENT OF CONTROL) OF INDIVIDUAL FACTORS AND THEIR
COMBINATION UPON MAIN ECOSYSTEM VARIABLES WITH FISH.  UPPER LINE FOR EACH CELL
REPRESENTS EXPERIMENTAL DATA, LOWER LINE = COMPUTER SIMULATION, SIGN. 1  =
SIGNIFICANCE LEVEL, INTER. = INTERACTION
Ecosystem
variables
Cladocera
abundance
Copepoda
abundance
Rotifer a
abundance
Algae
abundance
Bacteria
abundance
Number of
Cladocera
Copepoda
Rotifer a
Heavy
Effect
-30.03
-38.40
5.46
10.24
-57.28
-31.79
-21.82
1.04
10.12
1.03
species
-2.15
-1.29
-21.09
metals
Sign. 1
0.12
0.38
<0.01
0
<0.05
0.32

>0.50
>0.50
<0.01
Chlorpyrifos
Effect
-48.35
-46.42
31.06
23.66
13.98
-2.70
32.39
-1.62
-9.02
0.00

-28.04
-13.31
7.55
Sign. 1
<0.01
<0.05
<0.01
<0.01
0.39

<0.01
<0.05
0.19
Heavy metals
Effect
-73.62
-63.62
37.92
34.81
4.69
-33.63
3.40
-0.59
0.21
1.03

-30.57
-13.85
-15.56
Sign. 1
<0.01
<0.05
>0.50
0.43
>0.50

<0.01
0.06
0.07
4- chlorpyrifos
Inter.
4.75
21.19
1.39
0.91
-38.01
0.68
-7.16
-0.02
-0.89
0.00

0.39
0.75
-2.01
Sign. 1
0.37
>0.50
0.17
0.35
>0.50

-0.50
0.48
0.29
plankton in fishless habitats. Therefore, the indirect effects of fish in our experiments
consisted in the removal of invertebrate predators and large herbivores, thus securing the
proliferation of small planktonic organisms. The model also showed an increase of
protozoans which were not observed in the experiments.

       As far as the experimental design including treatments where more than one factor
was present at a time, it was possible to assess the combined effect of factors and to find out
if the factors interacted with each other. Interactions were calculated as differences between
observed effects of two or three factors present in a given treatment and the sum of the
individual effects of these factors taken from other treatments.  This sum was considered to
be equal to the additive effect.  Therefore, if the interaction effect coincides by its sign (plus
or minus) with the observed combined effect of factors, the interaction  is synergistic; if not, it
is antagonistic.  As seen from Tables 2 and 3, a combination of factors in most cases caused
purely  additive effects; interactions of factors were small in magnitude and statistically
insignificant.  Only some interactions involving fish and  observed through the effects on
rotifers and algae were large and significant, but these  are indirect secondary effects, so
these interactions may not be interactions of the factors themselves but just reflections of
interactions of the ecosystem living components.
                                         157

-------
       To see whether an ecosystem with one vertebrate predator on the top differs in
responses to toxicants from an ecosystem without such a predator, we divided the
experimental data into two, one with fish and another without fish, and performed the same
kind of analysis separately on each half. The results are presented in Tables 4 and 5.

       It can be seen from the tables that, in the presence of fish, effects of both toxicants on
crustaceans were damped down, the effect of heavy metals  on rotifers increased,  and the
effect of chlorpyrifos diminished.  It seems that there is no regularity in this pattern, but the
following reasoning may serve as a probable explanation. If a living population is well
developed and flourishing, then negative effects on it are pronounced, the positive ones are
barely observed, and vice versa. Cladocerans are well developed without fish, so adverse
effects of the toxicants were manifested better than with fish. Copepods attained high
numbers in the presence of fish; therefore, the stimulatory effect of toxicants on their
abundance was smoother.  Rotifers flourished in the presence of fish so the negative effect of
heavy metals on them was amplified, and the positive effect of chlorpyrifos was diminished.
Of course, these explanations should be taken cautiously because the fish presence imparted
great variability to experimental results and rendered many values statistically insignificant.
As for effects on algae and bacteria, at the present stage of the investigation they are hard to
interpret.

       When fish  are considered as another ecosystem component, the effects of toxicants
on their principal characteristics can be determined.  As seen in Fig. 1, heavy metals alone
and in combination with chlorpyrifos, each induced almost 2-fold  decreases in linear growth
and weight of fish. The greatest negative effect was caused by the combination of heavy
metals and the pesticide. In view of this, the significant stimulation of fish growth by 25%
compared to the control that was observed in the chlorpyrifos treatment is of great interest.

       Investigations of resistance of fish to waterflow,  reflecting  general physiological fitness
of fish (Pavlov and Fomin 1978), showed that the underyearlings which had been exposed to
chlorpyrifos  manifested the greatest resistance to flow (Fig. 2). These were, on the average,
18% more resistant than  those in the control.  In the heavy metals treatment, they showed
5% lower resistance, and in the treatment heavy metals + chlorpyrifos, their resistance was
reduced by 20%.  After 30 days of recovery from toxicants, the length and mass of fish
exposed to combined effects of chlorpyrifos and metals was greater than in control fish.  The
largest were fish from the chlorpyrifos treatment, although differences  in size and weight
were, on the whole, statistically insignificant.

       If we judged the whole ecosystem by fish conditions,  it would seem that the
ecosystem flourished when exposed to chlorpyrifos since the top link of the trophic chain was
in a better state than in clear water.  This misleading evidence shows that the whole
ecosystem testing involving many organisms  belonging to different trophic levels is more
appropriate to assess toxicant effects.
                                    CONCLUSIONS

       In general, results presented here agree well with those obtained by the
Environmental Research Laboratory, especially concerning major effects of toxicants on the
                                         158

-------
                66-
                80-
                56
                so •
                46 •
                40 •
                35 •
                33 •
                s-
                20
                     Ch    HM  Ch+HM  C
                          65
                          80-
                          JB
                          SO
                          45 •
                          40 •
                          S •
                          30 •
                          25 •
                          20 •
                                Ch   HM  Ch+HM   C
               MO
               220
               200
               180
               160
               140
               120
               103
               BO
                         240
                         220
                         200
                         180
                         180
                         140
                         120
                         103
                          60
                    Ch   HM  Ch-t-HM  C
                               Ch   HM   Ch+HM C
Fig. 1.  Growth in mass and length of bream underyearlingSs Experimental treatments: Ch =
chlorpyrifos, HM = heavy metals, Ch + HM = chlorpyrifos + heavy metals, C = control; a)
increases in length after 30 days of exposure, b) increases in length after 30 days of recovery, c)
increases in mass after 30 days of exposure, and d) increases in mass after 30 days of recovery.
                                 a                                     b
                9r
                    Ch
HM   Ch+HM
                                                           Ch    HM   Ch+HM   C
Fig. 2. Resistance of the fish to flow.  Index of resistance of fish to flow expressed as V/L, where
V is velocity of the flow (cm/sec) and L is length of the fish in sm; a) resistance of bream to flow
after 30 days of exposure and b) resistance of bream to flow after 30 days of recovery.
                                             159

-------
zooplankton community (U.S. EPA 1986).  Earlier, similar effects were shown in the studies
of Moscow University using static mesocosms (Filenko 1984).  Results demonstrate the
flow-through mesocosm protocol's utility to test biological effects of different kinds of toxicants
used separately or in combination.  In some aspects, the continuous system is better than
stationary systems, as it allows easier simulations of natural conditions, whereby pollutants
are usually being continuously added to a waterbody.  A good fit of the predicted and
observed effects of anthropogenous factors shows that mathematical modeling can be
successfully applied to such systems.
                                   REFERENCES

Anderson, R. L.  1982.  Toxicity of fenvalerate and permethrin to several nontarget aquatic
       invertebrates.  Environ. Entomol.  11:1251-1257.

Andrushkaitis, A. G., A. Kh. Avens, and Z. K. Seusuma.  1989.  The methods of
       ecotoxicological experiments in situ. Hydrobiol. J. 20:76-81.  (In Russian.)

Baudouin, M. F. and P. Scoppa. 1974.  Acute toxicity of heavy  metals to freshwater
       zooplankton. Bull. Environ. Contam. Toxicol.  12:745-751.

Buhr, H. O. and S. B. Miller.  1983. A dynamic model of the high-rate algal-bacterial
       wastewater treatment pond. Water Res.  17:29-37.

Dillon, W. R. and M. Goldstein.  1984. Multivanate analysis: methods and applications.  John
       Wiley and Sons, New York.

Eppley, R. W., J. N. Rogers and J. J. McCarthy.   1969.  Half-saturation constants for uptake
       of nitrate and ammonium by marine phytoplankton.  Limnol. Oceanogr. 14:912-920.

Filenko, O.  I.  (ed.). 1984.  Biological processes  in polluted model water-bodies.  M.,  MGU.
       (In Russian.)

Hurlbert, S. H., M. S.  Mulla and H. R. Willson. 1972.  Effects of organophosphorus
       insecticides on the phytoplankton, zooplankton, and insect populations of freshwater
       ponds. Ecol. Monogr. 42:269-299.

Hurlbert, S. H. 1975. Secondary effects of pesticides on aquatic ecosystems.  Residue Rev.
       57:81-148.

Kaushik, N. K., G. L. Stephenson,  K. A. Solomon and K. E. Day.  1985. Impact of permethrin
       on zooplankton communities in limnocorals.  Can. J. Fish. Aquat. Sci. 42:77-85.

Kokova, V.  E.  1982.  Continuous cultivation of fnvertebrates. Novosibirsk, Nauka SO.
       (In Russian.)

Kruychkova, N. M.  1974.  Food composition and the size of food particles consumed by
       planktonic fiber feeders. Hydrobiol. J. 10:117-123.  (In Russian.)
                                        160

-------
Lundgren, A. 1985.  Model ecosystems as a tool in freshwater and marine research. Arch.
       Hydrobiol. Suppl. 70:157-196.

Maly, M. and E. Ruber.  1983.  Effects of pesticides on pure and mixed species cultures of
       salt marsh pool algae. Bull. Environ. Contam. Toxicol. 30:464-472.

Moore, J. W. and S. Ramamoorthy.  1984.  Heavy metals in natural waters: applied
       monitoring and impact assessment.  Springer-Verlag, New York.  270pp.

Pavlov, D. S. and V. K. Fomin.  1978.  Methods of determination of critical flow rates for fish.
       Fisheries 11.  (In Russian.)

Roberts, D. and T. A.  Miller.  1970. The effects of diatoms on the larvicidal activity of
       Dursban.  November  1969-March  1970.  National Technical  Information Service,
       Springfield, VA.

Soto, D. and  S. H. Hurlbert.  1991.  Long-term experiments on calanoid-cyclopoid
       interactions. Ecol. Monogr. 61:245-265.

Stratton, G. W., A. L. Huber and C. T. Corke. 1980.  The effect of pesticides and their
       metabolites, alone and in combination, on algal processes.  Can. Tech. Rep., Fish.
       Aquat. Sci. 975:131-139.

U.S. Environmental Protection Agency. 1986.  Ambient water quality for chlorpyrifos - 1986.
       EPA 440/5-86/005, Washington, DC.

Vanni, M. J.  1986. Competition in the zooplankton communities: suppression of small
       species by Daphnia pulex.  Limnol. Oceanogr. 31:1039-1056.

Vanni, M. J.  1988. Freshwater zooplankton community structure: introduction of large
       invertebrate predators and large herbivores to a small species community.  Can. J.
       Fish. Aquat. Sci. 45:1758-1770.

Vinogradov, G. A. and V. B. Tagunov.  1989. An installation to study the effects of various
       substances on  fish and invertebrates under continuous flow conditions.  Hydrobiol. J.
       25.  (In Russian.)

Wylie, J.  L. and D. J. Currie.  1991. The relative importance of bacteria and algae as food
       sources for crustacean zooplankton.  Limnol. Oceanogr. 36:708-728.

Zaika, V. E. 1972. Specific production of invertebrates. Kiev, Naukova dumka. (In
       Russian.)

Zaika, V. E. 1983. Comparative productivity of hydrobionts.  Kiev, Naukova dumka.  (In
       Russian.)
                                        161

-------
          BIOCHEMICAL AND PHYSIOLOGICAL INDICATORS OF CONTAMINANT
             STRESS IN AQUATIC ORGANISMS OF LARGE RIVER SYSTEMS
                                          by

                          Denny R. Buckler1 and Donald E. Tillrtt1



                                      ABSTRACT

       Examination of the biochemical and physiological responses of organisms to xenobiotic
chemicals has been an area of scientific investigation for many years. Historically, these
techniques have been used in the development and screening of natural and synthetic
Pharmaceuticals, biocides, and other biologically active chemicals and in elucidating their modes
of action.  However, only recently have these techniques been applied as indicators  of the nature
and extent of chemical pollution of aquatic ecosystems. Today, research on the effects of
chemical contaminants at the molecular level of biological organization is one of the  most rapidly
developing and intensively studied areas in the field of environmental toxicology. In  this paper,
we discuss the relationship of biochemical and physiologic responses to effects at higher levels
of biological organization, examine general categories and provide specific examples of
techniques that are presently being applied to aquatic systems, and provide recommendations
for implementation of  biochemical and physiological techniques in aquatic contaminant
assessment programs.

                                   INTRODUCTION

       Water, soil, and sediments serve as the ultimate sinks for most chemicals produced and
used by man. General sources of anthropogenic chemical contaminants are from industrial,
agricultural, and urban uses. These uses result in both point and nonpoint source pollution of our
environment. Large rivers are of particular concern because they receive and integrate
pollutants from man's activities and land use practices over large geographical regions. Large
rivers are the direct recipients of municipal and industrial effluents and additionally receive
contaminant loadings from tributaries and agricultural practices in their watersheds.

       These multiple sources of contaminant input result in complex mixtures of contaminants
being present in the water and sediments of large river systems.  Unfortunately, most of our
knowledge and understanding of the effects of chemical contaminants in aquatic organisms is
based upon the effects of single compounds tested in the laboratory.  Much less is known about
the effects of complex environmental mixtures. Field studies can document the correlation  of
population-level effects with contaminant exposure, but the complexity of environmental variables
and multiple  contaminant exposure makes it difficult to establish cause and effect relationships
with specific  chemical pollutants. Environmental scientists are presently focusing more effort on
'National Fisheries Contaminant Research Center, U.S. Fish and Wildlife Service, Columbia, MO
65201-9643, USA


                                         162

-------
expanding our understanding of the effects of pollutants at the molecular level to help
discriminate among the various environmental stressors affecting aquatic populations and
communities.

       Management practices for aquatic resources and regulatory practices regarding
contaminant discharge are primarily directed at the population level. However, biochemical and
physiological measures of contaminant stress have received increased attention in recent years
due to their rapid response, sensitivity, and potential to provide a further level of discrimination in
the environmental risk assessment process.  In this paper, we discuss the relationship between
contaminant effects at the biochemical level and those at higher levels of biological organization,
examine general categories and specific examples of techniques that can be applied to aquatic
systems, and provide recommendations for integration of biochemical and physiological
measures of contaminant stress with other environmental risk assessment procedures.

             RELATIONSHIP OF BIOCHEMICAL AND PHYSIOLOGICAL STRESS
            TO EFFECTS AT HIGHER LEVELS OF BIOLOGICAL ORGANIZATION

       A basic premise in the field of toxicology is that to elicit a toxic response, a chemical must
first reach and interact with its biomolecular site of action.  This response on a molecular level
may then be followed by a tissue response, an organ response, a whole animal response, and
finally a population or community response,  but it must occur first at the biomolecular level.

       Each level possesses an inherent resiliency, or assimilative capacity, that allows it to
maintain homeostasis. To provide a continuum of response, the magnitude of effect at each
level of organization must be great enough to exceed the assimilative capacity  of the next higher
level. Ultimately, for a contaminant to elicit a response at the population or community level, it
must exceed the capacity of all  lower levels to absorb  or mitigate the insult.

       These relationships and interdependencies provide both promise and challenge to the
use of biochemical and physiological responses as indicators of aquatic community health. First,
it is apparent that biochemical changes must be the first and most sensitive biological responses
that occur following contaminant exposure. Thus, bioindicators may be useful as early warning
indicators of contaminant problems. Alternatively, because the continuum  may be broken at
subsequent levels of biological organization, effects at the  biochemical level may not be
ecologically significant.  Thus, the challenge facing the environmental scientist is to develop an
understanding of the magnitude of effect at the biochemical level  that is required to elicit an
ecologically significant response at higher levels of organization.

       Toxicants interact with organisms on the molecular level in a variety of ways that can
challenge the maintenance of homeostasis.  Metals can bind to various proteins disrupting
membrane integrity, ion transport, and cellular metabolism.  For example, mercury can bind to
sulfhydryl groups on structural proteins (Luckey and Venugopal 1977), thereby  disrupting
membrane integrity. Copper can interfere with ion transport by affecting  gill ATPase activity,
reducing the ability of coho salmon (Oncorhynchus kisutch) to adapt to seawater (Lorz and
McPherson 1976).  Lead can deactivate 6-aminolevulinic acid dehydratase (ALAD), disrupting
red blood cell metabolism.

       Organic contaminants also interact with organisms on the molecular level by disrupting
membrane integrity and function, displacing endogenous substrates of enzymes, or by
diminishing energy reserves through increased demands upon detoxification mechanisms. For


                                         163

-------
example, pentachlorophenol uncouples oxidative phosphorylation by interacting with the
mitochondrial membrane (Moreland 1980). Organophosphates and carbamates can displace
acetylene-line and inhibit its hydrolytic cleavage by acetylcholinesterase (ACHe), thereby affecting
nerve transmission (O'Brien 1976). Widdows and Donkin (1991) report a variety of ways that
toxicants can diminish the energy reserves of an organism to the detriment of survival.

       To cause effects at higher levels of biological organization, the effects at the molecular
level mentioned above must be of sufficient severity to challenge the organism's ability to
maintain homeostasis. Organisms invite a variety of mechanisms for dealing with toxicants. In
some cases, they may be able to detect and avoid the toxicant (Folmar 1976).  They may
sequester the toxicant (e.g., through induction of metal-binding proteins, such as
metallothioneins), or they may metabolically alter the compound to a less toxic form that may be
more readily eliminated (e.g., oxidative metabolism of pyrethroid insecticides).  Alternatively, they
may simply repair the cellular damage caused by the toxicant. Each of these mechanisms incurs
some physiologic cost on the  part of the individual organism (Calow 1991). When the organism's
ability to absorb or mitigate the toxic insult is exceeded, stress will be applied to the higher levels
of organization such as the population or community.

       At present, a wide range of biochemical and  physiological parameters are known to be
responsive to chemical exposure. However, our understanding of how these responses are
quantitatively linked to effects at the aquatic population and community levels is much more
limited.  Relatively few of the  measurable biochemical and physiological responses are specific to
certain chemical classes (e.g., ACHe inhibition by Organophosphates and carbamates; ALAD
inhibition by lead).  The majority of responses  are sensitive to a wide variety of  chemical
compounds and other environmental stressors. Thus, with a few exceptions, the majority of
biochemical and physiological measures are more useful as general indicators of organism
health rather than specific indicators of specific chemical exposure. It is important to recognize
the range of sensitivity of the  indicators we use in aquatic contaminant risk assessment, because
our goal is to establish cause and effect between contaminant exposure and  responses at the
population and community level. Responses that are general indicators of organism health must
be used in concert with other  measures of chemical  exposure, such as chemical residue
analysis. Responses that are more specific indicators of individual chemical exposure must be
calibrated to quantify the continuum of response from the biochemical to the population and
community level.

       Within the broad range of measurable biochemical and physiological parameters, a
further distinction can be made between regulated parameters and regulatory mechanisms.
Regulated parameters are usually maintained within relatively narrow ranges to maintain
physiological function. Contaminant exposure that displaces these parameters beyond the
tolerated range will usually result in acute injury or death (Zachariassen et al. 1991).  However,
the regulatory mechanisms that govern  these parameters may fluctuate much more widely before
significant injury or death occurs in an attempt to maintain the optimal range, and may be more
responsive and  sensitive indicators of contaminant insult (Zachariassen et al. 1991). For
example, the level of ATP present in an organism is  a highly regulated parameter.  ATP levels
can be maintained at the expense of other endogenous high-energy phosphate compounds.
Aunaas et al. (1991) demonstrated that sublethal exposure of the blue mussel (Mytilus edulis) to
contaminants generally resulted in larger fluctuations of components of the phosphate pool other
than ATP. Thus, the regulatory mechanism (the phosphate pool) was more responsive than the
regulated parameter (ATP).
                                          164

-------
       The range of sensitivity, level of variability, and relationship to responses at higher levels
of biological organization are all important parameters for consideration in selecting biochemical
and physiological indicators for application in aquatic contaminant risk assessment programs.

                     CATEGORIES AND EXAMPLES OF TECHNIQUES

       Biochemical and physiological indicators of contaminant stress can be categorized in a
variety of ways-general versus specific sensitivity to compounds, regulatory versus regulated
parameters, indicators of exposure versus indicators of effect, or by category of biochemical and
physiological function.  In this section, we will identify general categories of biochemical and
physiological function and discuss representative examples of indicator techniques for each
category that are presently being  used or show promise for further development.

OSMOREGULATION

       A variety of inorganic and  organic environmental pollutants affect osmoregulation in
fishes. For this reason, many osmoregulatory stress indicators are more useful as indicators of
general organism health rather than diagnostic tools for identification of specific pollutants.

       Plasma ion concentrations are responsive to heavy rnetal exposure. Sublethal exposure
to copper causes transient decreases in plasma chloride concentrations in brown bullheads
(Ictalurus nebulosus) and brook trout (Satvelinus fontinalis)', however, the levels return to normal
after about 3 weeks even during continuous exposure,  indicating the presence of a
compensatory mechanism for maintenance of homeostasis (Christensen et al.  1972, McKim et
al, 1970). Zinc and cadmium also affect plasma ion concentrations (Lewis and Lewis 1971,
McCarty  and Houston 1976).

       Gill ATPase activity in fishes is also sensitive to chemical exposure. This measure of
osmoregulatory ability is affected  by a wide range of environmental pollutants including copper
(Lorz and McPherson 1976), PCBs (Davis et al.  1972), and a variety of chlorinated pesticides
(Davis et al. 1972).  Gill ATPase activity has direct ecological relevance in that it is intimately
linked to  the migratory success of anadromous fish species (Epstein et al. 1967, Zaugg and
Wagner 1973).

       Histological and histochemical examination of gill tissue has also been used successfully
to demonstrate the effects  of osmoregulatory stressors. As with ATPase activity, a wide variety
of environmental contaminants can cause histological lesions in fish gill.  Aluminum (Jagoe et al.
1987), copper  (Baker 1969), mercury (Wobeser 1975),  sodium arsenite (Gilderhaus 1966), and
zinc (Skidmore and Tovell 1972) are some of the metals that cause gill lesions  in fishes. A wide
variety of organic compounds also cause histological lesions in fish gill.  Endrin (Eller 1971),
heptachlor  (Andrews et al.  1966),  methoxychlor (Lakota et al 1978), mirex (Van Valin et al. 1968),
naphthalene (DiMichele and Taylor 1978), and phenol (Mitrovic et al. 1968) are only a few of the
organic contaminants that cause gill lesions.

METAL SEQUESTRATION AND REGULATION

      Metallothioneins are a class of metal-binding proteins found in a wide variety of
organisms (Roesijadi 1992) that are inducible after exposure to Cd, Cu, Hg, and Zn (Noel-Lambot
et al. 1978). Metallothioneins are  thought to reduce the toxicity of heavy metal  exposure by
sequestration (Hamilton and Mehrle 1986), with toxicity occurring only after its binding capacity  is


                                          165

-------
exceeded (Brown and parsons 1978). Hepatic metallothionein levels in feral fishes from lakes
contaminated with Cu, Zn and Cd are closely correlated to metals exposure (Roch et al. 1982).
The inducible nature of metallothioneins suggests that they represent a regulatory mechanism,
and as such provide a useful indicator of exposure to certain metals.  Other metal-binding
proteins such as ferritin (Fe) and copper-chelatin (Cu) have been less thoroughly studied in
aquatic organisms.

OXIDATIVE METABOLISM

       Oxidative metabolism serves three important functions in eukaryotic organisms that
provide useful biochemical indicators of stress.  First, oxidative metabolism plays a central role in
catabolic energy production.  The resultant storage currency is adenylate, and its utility as a
biochemical indicator of contaminant stress is discussed in the next section.

       Secondly, oxidative metabolism is associated with the reduction of oxygen radicals and
the maintenance of  oxidative homeostasis within the cell. This critical function also lends itself
for use as a biochemical indicator of stress in aquatic organisms.  Oxygen  radicals are formed as
intermediates of cytosolic enzymes, electron transfer systems, and by activated cells of the
immune system.  Inducible prophylactic enzymes such as superoxide dismutase, catalase, and
glutathione peroxidase serve a vital role in protecting the cell from oxidative stress and are useful
indicators of contaminant stress in aquatic organisms (DiGiulio et al. 1989). Additionally, the
pools (and relative ratios) of reducing equivalents (e.g., glutathione, a-tocopherol, and
ascorbate) that  are required to maintain oxidative homeostasis are useful indicators of
contaminant stress.  These antioxidant defense mechanisms of the cell are sensitive to
compounds that block or inhibit electron flow and, in particular, those that can undergo cyclic
reactions (redox cycling) to produce oxygen radicals. These important classes of environmental
contaminants include paraquat, diquat, and various amines.

       The third, and probably the most thoroughly studied, function of oxidative metabolism is
xenobiotic metabolism. Xenobiotic metabolism associated with the cytochrome P450
monooxygenases (MO) has been reviewed for use in biomonitoring by Payne et al. (1987), who
concluded that induction of MO could be a sensitive biochemical indicator of exposure in  many
aquatic organisms.  Induction of MO activity is caused by a wide variety of chemical
contaminants (e.g.,  PAHs, PCBs,  PCDDs, and PCDFs). The MO system plays a central role in
the detoxification of  these and many other hydrophobic compounds largely by facilitating  their
elimination.  Additionally, MOs are highly conserved phylogenetically. All of these characteristics
enhance the potential utility of MO activity as a biochemical indicator of contaminant stress.
However, its use in feral organisms has been limited because of the confounding factors  of age,
sex, reproductive status, diet, disease, and general health conditions that can all greatly affect
the MO response to xenobiotics (Neal 1980). A thorough understanding of the effects caused by
these variables  is lacking in most species of fish and wildlife.  Even if these relationships  were
understood in aquatic species, it is often difficult to obtain such information on feral organisms.
Thus, MO activity in this context offers a more qualitative than quantitative measure of
contaminant exposure. However, successful use of MO activity as a quantitative bioindicator of
contaminant stress in the field has been accomplished using caged fish studies (Lindstrom-
Seppa and Oikari 1990) and by using developing embryos (Hoffman et al. 1987). In both of
these cases, the organismal and environmental factors affecting MO activity could be controlled
or monitored during  the period of exposure.
                                          166

-------
       Another alternative to limiting the confounding factors that influence MO activity is the use
of in vitro systems where the conditions are tightly controlled. In these assays, the
environmental contaminant mixture is brought to the laboratory to expose the system, as
opposed to taking the system (organisms) to the environment for exposure.  An example of such
an in vitro model system for MO activity is the H4IIE hepatoma cell bioassay (Tillitt et al. 1991).
In this bioassay, environmental contaminants are extracted from the species of concern and
serial dilutions are used to dose the H4IIE cells. After a period of incubation, ethoxyresorufin-o-
deethylase (EROD) activity is measured in the cells. The MO response of the H4IIE cells is
highly correlated to the effects observed in the whole organism (Safe 1987). Therefore, the MO
response in vitro can integrate the interactions of complex mixtures of environmental
contaminants at the cellular/biochemical level and serve as a quantitative measure of both
exposure and effect in the whole organism from which the contaminants were extracted. Recent
studies have demonstrated the ability of the H4IIE bioassay to be predictive of contaminant
responses at higher levels of biological organization, such as egg mortality rates in PCB-exposed
populations of fish-eating birds (Tillitt et al. 1992).

MAINTENANCE OF ENERGY STATUS

       Adenylate energy charge is a term developed by Atkinson (1968) to describe the dynamic
equilibrium between the various components of the cellular energy system-ATP, ADP, and AMP.
It is calculated as (ATP + 0.5 X ADP)/(ATP + ADP + AMP) and its value reflects the energy
status of the organism. Adenylate energy charge is sensitive to a variety of environmental
factors such as food availability and temperature (Moal et al. 1991), season (Giesy and Dickson
1981), salinity  (Ivanovici 1980), and pollutants (Verschraegen et al.  1985). Because ATP  levels
are regulated within relatively narrow limits in many organisms, Aunaas et al. (1991) have
proposed the use of a phosphorus index-which reflects the status of a larger phosphorus pool--
as a more sensitive indicator of energy status.  The broad  spectrum of environmental stressors
to which energy status is sensitive indicates that this biochemical parameter may be most useful
as a general indicator of organism health, rather than a specific indicator of contaminant
exposure.

REPRODUCTION

       Another category of biochemical and physiological  function that is sensitive to
contaminant exposure is reproduction. The complexity of the reproductive process in aquatic
organisms and the confounding influence of behavior, nutritional status, seasonality, and other
variables makes it difficult to ascertain the effect of contaminants on the reproductive success of
natural populations. However, a number of biochemical reproductive parameters have been
investigated in the laboratory in terms of their sensitivity to contaminant exposure. Vitellogenin  is
the major yolk  protein in salmonids and many other species. It is produced in the liver and
transported to the ovaries in the blood. Blood levels of vitellogenin  are elevated during ovary
maturation.  Vitellogenin levels in brook trout and rainbow trout are affected by low pH exposure
(Tam et al.  1987, Parker and McKeown 1987).  Reproductive endocrine function in fishes has
also been studied to a limited extent in terms of the effects of contaminants on steroid hormone
levels. Thomas (1990) demonstrated  an increase in plasma hormone levels in Atlantic croaker
(Micropogonias undulatus) after exposure to lead, benzo(a)pyrene, or Arochlor 1254;
vitellogenin levels were similarly affected by cadmium and  PCBs.
                                          167

-------
NEUROTRANSMISSIOM

       One of the more specific biochemical indicators of contaminant exposure is measurement
of the neurotransmitter acetylcholinesterase (ACHe). Organophosphates and carpamrates are
specific inhibitors of ACHe activity (O'Brien 1976), although other physiological factors can
influence activity levels. It should be noted that measurement of brain activity levels represents
primarily true acetylcholinesterase whereas plasma levels generally reflect both
acetylcholinesterase and butyryl(pseudo)cholinesterase (O'Brien 1976).  Depressed ACHe
activity is associated with a variety of behavioral responses in fishes (Symons 1973) and it has
been measured in feral fishes from polluted waters (Williams and Sova 1966). However, brain
activity levels  resulting in death of fishes can be quite variable (Weiss 1961).

INTERACTIONS WITH GENETIC MATERIAL

       The genotoxic effects of contaminants are receiving increased interest in the area of
environmental contaminant assessment.  A variety of techniques are becoming available for use
in aquatic systems. Detection of the presence of DNA adducts using 32p-postlabeling is one
method of assessing exposure to contaminants that are capable of interacting with genetic
material.  Varanasi et al. (1989) used this method to correlate  DNA adducts in livers of English
sole (Parophrys vetulus) and winter flounder (Pseudopleuronectes americanus from Pugent
Sound, with exposure to PAHs in sediments. Another method that shows promise is the use of
the DNA alkaline unwinding assay.  Toxic chemicals that interact with genetic material can cause
strand breaks in DNA. The rate of unwinding of the DNA double helix when placed in an alkaline
solution is a measure of DNA integrity that has been shown to be related to exposure of fish to
genotoxic chemicals (Shugart 1988).

IMMUNOLOGY

       A wide range of environmental pollutants affect the immune system of fishes (Zeeman
and Brindley 1981).  Immunosuppression can result in increased susceptibility of fishes to
disease (Sinderman  1979). Contaminant-reduced immunocompetence of fishes has been a
subject of study in both feral fishes from polluted environments (Warinner et al. 1988), and
laboratory studies with single compounds (Walczak et al. 1987) and complex mixtures
(Secombes et al. 1991). Various approaches to assessment of immune status have been used.
These include differential white blood cell counts (Gardner and Yevich 1970), serum protein
levels and ratios (Smith et al. 1976), agglutination liter (Roales and Perlmutter 1977),
phagocytosis (Secombes et al. 1991), and disease susceptibility (Hansen et al. 1971).  As with
many other biochemical and physiological indicators, measures of immunocompetence are more
useful as indicators of general organism health rather than a diagnostic indicator of specific
contaminant exposure.

                      RECOMMENDATIONS FOR IMPLEMENTATION

       In this  brief overview, we have had the opportunity to focus on only a minute fraction of the
biochemical and physiological response of aquatic organisms that have been measured in relation
to contaminant exposure.  Of the literally hundreds of measurable biochemical and physiological
responses of organisms, most have been studied at least to some extent in relation to exposure to
one chemical or another. The wide variety of species of aquatic organisms and the wide variety of
chemicals that pose a threat as environmental pollutants provides an extremely extensive and
complex matrix of possible combinations of exposure and response  for study.


                                         168

-------
       Relatively few biochemical and physiological responses of aquatic organisms are specific
to individual chemicals (e.g., ALAD); several others are sensitive to general chemical classes
(e.g., acetylcholinesterase activity; EROD activity). These specific indicators may be very useful
in environmental assessments where sources and types of contaminants are unknown, by
directing evaluation efforts toward particular chemicals. The sensitivity of these tests may also
make them useful for documenting the fringes of environmental disturbance or areas of subtle
contaminant exposure when the types of contaminants present in a system are known. Their
use, coupled with supportive analytical chemistry, can provide important information for
establishing cause and effect relationships and interpreting observed effects at higher levels of
biological organization.

       The majority of biochemical and physiological responses of aquatic organisms are
sensitive to broad ranges of chemical contaminants and other environmental stressors.  For this
reason, these techniques can provide sensitive measures of organism health, but are not
diagnostic for specific chemical exposure. These general indicators are most effectively applied
in environmental contaminant assessments in concert with other measures of contaminant
effects and exposure. The use of general biochemical indicators should be coupled with
chemical analysis, on-site and in-situ toxicity assays, infaunal surveys, and other classical
contaminant assessment procedures to identify the nature and extent of the contaminant
problem. As with specific indicators of contaminant exposure, these general indicators of
organism health may provide additional sensitivity in identifying the fringes of environmental
disturbance and areas of subtle contaminant exposure. However, supporting information for
establishing cause and effect is extremely important when general indicator techniques are
applied in environmental contaminant assessments.

       In summary, measurement of the biochemical and physiological responses of aquatic
organisms to toxic chemicals has provided extensive insight into the modes of action of
environmental contaminants. However, their use in the identification and monitoring of the nature
and extent of aquatic pollution is a relatively new application. Their rapid response and
sensitivity provides potential for adding a further level of discrimination in the environmental risk
assessment process, particularly in marginally impacted areas. Proper application of these
techniques requires a thorough understanding of their specificity in regard to the types of
environmental stressors to which they are sensitive,  and a quantitative understanding of the
linkages between effects at various levels of biological organization. For many biochemical and
physiological indicators, these are topics  of current and future work.

                                      REFERENCES

Aunaas, T., S. Einarson, T.E. Southon, and K.E. Zachariassen. 1991. The effects of organic
       and inorganic pollutants on intracellular phosphate compounds in blue mussels (Mytilus
       edulis).  Comp. Biochem. Physiol. 100C:89-93.

Andrews, A.K., C.C. Van Valin, and B.E. Stebbings.  1966. Some effects of heptachlor on
       bluegills (Lepomis macrochirus). Trans. Am. Fish. Soc. 95:297-309.

Atkinson, D.E. 1968. The energy charge of the adenylate pool as a regulatory parameter.
       Interaction with feedback modifiers. Biochemistry 7:4030-4034.
                                          169

-------
Baker, J.T.P. 1969. Histological and electron microscopical observations on copper poisoning in
       the winter flounder (Pseudopleuronectes americus). J. Fish Res. Board Can. 26:2785-2793.

Brown, D.A. and T.R. Parsons. 1978. Relationship between cytoplasmic distribution of mercury
       and toxic effects to zooplankton and chum salmon, Oncorhyncus keta, exposed to
       mercury in a controlled ecosystem. J. Fish. Res. Board Can. 35:880-884.

Calow, P. 1991. Physiological costs of combating chemical toxicants: Ecological implications.
       Comp. Biochem. Physiol. 100C:3-6.

Christensen, G.M., J.M. McKim, W.A. Brungs, and E.P. Hunt.  1972. Changes in the blood of the
       brown bull head (Ictalurus nebulosus [Leseur]) following short and long term exposure to
       copper(ll). Toxicol. Appl. Pharmacol. 23:417-427.

Davis, P.W., J.M. Friedhoff, and G.A. Wedemeyer.  1972. Organochlorine insecticide, herbicide
       and polychlorinated btphenyl (PCB) inhibition of Na K-ATPase in rainbow trout. Bull.
       Environ. Contam. Toxicol. 8:69-72.

DiGiulio,  R.T., P.C. Washburn, R.J. Wenning, G.W. Winston, and C.S. Jewel. 1989. Biochemical
       responses in aquatic animals:  A review of determinants of oxidative stress. Environ.
       Toxicol. Chem. 8:1103-1123.

DiMichele, L. and M.H. Taylor. 1978.  Histopathological and physiological responses of Fundulus
       heteroclituslo naphthalene exposure. J. Fish. Res. Board Can. 35:1060-1066.

Eller,  L.L. 1971. Histopathologic lesions in cutthroat trout (Salmo clarki) exposed chronically to
       the insecticide endrin. Am. J. Pathol. 64:321.

Epstein, F.H., A.I. Katz, and  G.E. Pickford. 1967. Sodium and potassium-activated adenosine
       triphosphatase of gills: Role in adaptation to teleosts to salt water. Science 156:1245-
       1247.

Folmar, L.C.  1976. Overt avoidance reaction of rainbow trout fry to nine herbicides. Bull.
       Environ. Contam. Toxicol. 15:509-514.

Gardner, G.R. and P.P. Yevich. 1970.  Histological and hematological responses of an estuarine
       teleost to cadmium. J. Fish. Res. Board Can. 27:2185-2196.

Giesy, J.P.  and G.W. Dickson. 1981.  The effects of season and location on phosphoadenylate
       concentrations and adenylate energy charge in two species of freshwater clams.
       Oecologia49:1-7.

Gilderhaus, P.A. 1066. Some effects of sublethal concentrations of sodium arsenite on bluegills
       and the aquatic environment.  Trans. Am. Fish. Soc. 95:289-296.

Hamilton, S.J. and P. M. Mehrle. 1986. Metallothionein in fish: Review of its importance in
       assessing stress from metal contaminants.  Trans. Am. Fish. Soc. 115:569-609.
                                          170

-------
Hansen, D.J., P.R. Parrish, J.I. Lowe, A.J. Wilson, Jr., and P.O. Wilson. 1971. Chronic toxicity,
       uptake and retention of Aroclor 1254 in two estuarine fishes. Bull. Environ. Contam.
       Toxicol. 6:113-119.

Hoffman, D.J., B.A. Rattner, L. Sileo, D. Docherty, and T. Kubiak.  1987. Embryotoxicity,
       teratogenicity, and aryl hydrocarbon hydroxylase activity in Forster's terns in Green Bay,
       Lake Michigan.  Enivron. Res. 42:176-184.

Ivanovici, A.M.  1980. The adenylate energy charge in the estuarine mollusc, Pyrazus ebeninus.
       Laboratory studies of responses to salinity and temperature. Comp. Biochem. Physiol.
       66A:43-55.

Jagoe, C.H., T.A. Haines, and D.R. Buckler. 1987. Abnormal gill development in Atlantic
       Salmon (Salmo salai) fry exposed to aluminum at low pH. Annls. Soc. R. Zool. Belg.
       117:375-386.

Lakota, S., A. Raszka, I. Kupczak, S.  Hlond, J. Stefan, and J. Roszkowski.  1978. The effect of
       methoxychlor and propoxur on the health of carp fry (Cyprinus carpio L).  Acta Hydrobiol.
       20:197-205.

Lewis, S.D. and W.M. Lewis.  1971.  Effect of zinc and copper on the osmolarity of blood serum
       of the channel catfish, Ictalurus punctatus Rafinesque, and the golden shiner,
       Notemigonus crysoleucas Mitchill.  Trans. Am. Fish. Soc.  100:639-643.

Lindstrom-Seppa, P.  and A. Oikari. 1980. Biotransformation and other toxicological and
       physiological responses in  rainbow trout  (Salmo gairdneri Richardson) caged in a lake
       receiving pulp and paper mill effluents. Aquat. Toxicol. 16:187-204.

Lorz, H.W.  and B.P. McPherson.  1976.  Effects of copper or zinc in fish on the adaptation to sea
       water and ATPase activity and the effects of copper on migratory disposition of coho
       salmon (Oncorhyncus kisutch). J. Fish. Res. Board Can.  33:2023-2030.

Luckey, T.D. and B. Venugopal. 1977. Metal Toxicity in Mammals. Plenum Press.

McCarty, L.S. and A.H. Houston.  1976.  Effects of exposure to sublethal levels of cadmium upon
       water-electrolyte status in the goldfish (Carassius auratus. J. Fish. Biol. 9:11-19.

McKim, J.M., G.M. Christensen, and E.P. Hunt.  1970. Changes in  the blood of brook trout
       (Salvelinus fontinalis) after  short-term and long-term exposure to copper.  J. Fish. Res.
       Board Can. 27:1883-1889.

Mitrovic, V.V., V.M. Brown, D.G. Shurben, and M.H. Berryman. 1968. Some pathological effects
       of subacute and acute poisoning of rainbow trout by phenol in hard water.  Water Res.
       2:249-254.

Moal, J., J.R. LeCoz,  J.F. Samain,  and J.Y. Daniel. 1991. Adenylate energy charge: A possible
       trophic index for management of oyster intensive aquaculture. Comp. Biochem. Physiol.
       100C:201-205.
                                          171

-------
Moreland, D.E.  1980. Effects of toxicants on oxidative and photophosphorylation.  In:
       Introduction to Biochemical Toxicology, E. Hodgson and F.E. Guthrie, eds. Elsevier. pp.
       245-260.

Neal, R.A. 1980. Metabolism of toxic substances. In: Casarett and Doull's Toxicology the Basic
       Science of Poisons, J. Doull, C.D. Klaassen, and M.O. Amdur, eds.  McMillan, pp. 56-69.

Noel-Lambot, F., C. Gerday, and A. Disteche.  1978. Distribution of Cd, Zn and Cu in liver and
       gills of the eel, Anguilla anguilla, with special reference to metallothioneins.  Comp.
       Biochem. Physiol. 610:177-187.

O'Brien, R.D. 1976. Acetylcholinesterase and its inhibition. In:  Insecticide Biochemistry and
       Physiology, C.F. Wilkinson, ed.  Plenum Press,  pp. 271-296.

Parker, D.B. and B.A. McKeown.  1987.  Effects of pH and/or calcium-enriched freshwater on
       plasma levels of vitellogenin and Ca2+ and on bone Ca content during exogenous
       vitellogenesis in rainbow trout (Salmogairdneri).  Comp. Biochem. Physiol. 87A:267-273.

Payne, J.E., L.L. Fancey, A.D. Rabimtula, and E.L. Porter. 1987. Review and perspective on
       the use of mixed-function oxidase enzymes in biological monitoring.  Comp. Biochem.
       Physiol.  86C:233-245.

Roales, R.R. and A.  Perlmutter. 1977.  The effects of sublethal doses of methylmercury and
       copper, applied singly and jointly, on the immune response of the blue gourami
       (Trichogaster trichopterus) to viral and bacterial antigens. Arch. Environ. Contam.
       Toxicol.  5:325-331.

Roch, M., J.A. McCarter, AT. Matheson, M.J.R. Clark, and R.W. Olafson. 1982.  Hepatic
       metallothionein in rainbow trout (Salmo gairdneri) as an indicator of metal pollution in the
       Campbell River System. Can. J. Fish. Aquat. Sci.  39:1596-1601.

Roesijadi, G. 1992.  Metallothioneins in metal regulation and toxicity in aquatic animals.  Aquat.
       Toxicol.  22:81-115.

Safe, S.  1987.  Determination of 2,3,7,8-TCDD toxic equivalent factors (TEFs): Support for the
       use of the in vitro AHH induction assay.  Chemosphere 16:791-802.

Secombes, C.J., T.C. Fletcher, J.A. O'Flynn, MJ. Costello, R. Stagg, and D.F. Houlihan.  1991.
       Immunocompetence as a measure of the biological effects of sewage sludge pollution in
       fish.  Comp. Biochem. Physiol. 100C:133-136.

Shugart,  L.R. 1988. Quantification of chemically induced damage to DMA of aquatic organisms
       by alkaline unwinding assay. Aquat. Toxicol. 13:43-52.

Sinderman, C.J. 1979. Pollution-associated diseases and abnormalities of fish and shellfish: A
       review. Fish. Bull. 76:717-749.

Skidmore, J.F. and P.W.A. Tovell. 1972. Toxic effects of zinc sulfate on the gills of rainbow trout.
       Water Res.  6:217-230.
                                          172

-------
Smith, B.P., E. Hejtmancik, and B.J. Camp. 1976.  Acute effects of cadmium on Ictalurus
       punctatus (catfish). Bull. Environ. Contam. Toxicol. 15:271-277.

Symons, P.E.K.  1973. Behavior of young Atlantic salmon exposed to or force-fed fenitrothion,
       an organophosphate insecticide. J. Fish. Res. Board of Can.  30:651-655.

Tarn, W.H., L. Birkett, R. Makaran, P.O. Payson, O.K. Whitney, and C.K.-C. Yu.  1987.
       Modification of carbohydrate metabolism and liver vitellogenin function in brook trout
       (Salvelinus fontinalis) by exposure to low pH. Can. J. Fish. Aquat. Sci. 44:630-635.

Thomas, P.  1990. Teleost model for studying the effects of chemicals on female reproductive
       endocrine function. J. Exp. Zool. 4:126-128.

Tillitt, D.E., J.P. Giesy, and G.T. Ankley.  1991.  Characterization of the H4IIE rat hepatoma cell
       bioassay as a tool for assessing toxic potency of planar halogenated hydrocarbons
       (PHHs) in environmental samples. Environ. Sci. Technol. 25:87-92.

Tillitt, D.E., G.T. Ankley, J.P. Giesy, J.P. Ludwig, H. Kurita-Matsuba, D.V. Weseloh, P.S. Ross,
       C.A. Bishop, L. Sileo, K.L. Stromborg, L. Larson, and T.J. Kubiak.  1992. Polychlorinated
       biphenyl residues and egg mortality in double-crested cormorants from the Great Lakes.
       Environ. Toxicol. Chem. 11:1281-1288.

Van Valin, C.C., A.K. Andrews, and L.L. Eller.  1968. Some effects of mirex on two warm-water
       fishes. Trans. Am. Fish. Soc. 97:185-196.

Varanasi, U., W.L. Reichert, and J. Stein. 1989. 32P-postlabeling analysis of DNA adducts in
       liver of wild English sole (Parophyrs vetulus) and winter flounder (Pseudopleuronectes
       americanus).  Cancer Res. 49:1171-1177.

Verschraegen, K., P.M.J. Herman, D. Van Gansbeke, and A. Braeckman.  1985. Measurement
       of the adenylate energy charge  in  Nereis diversicolor and Nephtys sp. (Polichaeta:
       Annelida). Evaluation of the usefulness of AEC in pollution monitoring. Mar. Biol.
       86:233-240.

Walczak, B.Z., B.R. Blunt, and P.V.  Hodson. 1987. Phagocytic function of monocytes and
       haematological changes in rainbow trout injected intraperitoneally with benzo(a)pyrene
       [B(a)P] and benzo(a)anthracene [B(a)A]. J. Fish. Biol. 31 A-.251-253.

Warinner, J.E., E.S. Mathews, and B.A. Weeks.  1988.  Preliminary investigations of the
       chemiluminescent response in normal and pollutant-exposed fish. Mar. Environ. Res.
       24:281-284.
Weiss, C.M. 1961.  Physiological effect of organic phosphorus insecticides on several species of
       fish.  Trans. Am. Fish. Soc. 90:143-152.

Widdows, J. and P. Donkin. 1991. Role of physiological energetics in ecotoxicology. Comp.
       Biochem. Physiol.  100C:69-75.
                                          173

-------
Williams, A.K. and R.C. Sova.  1966.  Acetylcholinesterase levels in brains of fishes from polluted
       waters. Bull. Environ. Contam. Toxicof. 1:198-204.

Wobeser, G. 1975.  Acute toxicity of methyl mercury chloride for rainbow trout (Salmo gairdneri)
       fry and  fingerlings. J. Fish. Res. Board Can. 32:2005-2013.

Zachariassen,  K.E., T. Aunaas, J.F. Borseth, S. Einarson, T. Nordtug, A. Olsen, and G. Skjero.
       1991.  Physiological parameters in ecotoxicology. Comp. Biochem. Physiol. 100C:77-79.

Zaugg, W.S. and H.H. Wagner. 1973. Gill ATPase activity related to par-smolt transformation
       and migration in steelhead trout (Salmo gairdneri): Influence of photoperiod and
       temperature. Comp. Biochem. Physiol. 45:955-965.

Zeeman, M.G.  and W.A. Brindley. 1981. Effects of toxic agents upon fish immune systems: A
       review. In: Immunologic Considerations in Toxicology, Vol. II, R.P. Sharma, ed. CRC
       Press, pp. 1-60.
                                         174

-------
                     CLASSIFICATION OF WETLANDS OF RUSSIA

                                          by

                    A. M. Nikanorov1, A. V. Zhulidov1 and W. Sanville2


                                     ABSTRACT

       The term "wetland," which is new for the Russian scientific literature, describes a
diversity of zones lying between typical terrestrial habitats and deep water ones.  There is no
unified classification of wet habitats in Russia. In order to classify Russian wetlands, we
used the well-known publication, "Classification of Wetlands in the United States," because of
its ecological comprehensiveness and simplicity.


                                   INTRODUCTION

       Habitats at the water and land interface are complex structures in which we observe
the interaction of water, air, soil, bottom sediments, and living organisms.  These habitats are
quite common at marine coasts, lake and river shores, relief depressions, and  ground water
outlets.  In Russia we  use a whole set  of terms designating such ecosystems.  They may be
terms which are quite  common, such as wet meadows, bogs, swamps, and coasts or local
specific terms such as laid, alas, plav, tugai, zaimizche, kardash, iskar, urem, grass visk,
sogr, saz, ryam, ols, vatt, badaran,  etc. Sometimes, one term is understood differently in
different areas of Russia.  For instance, the term "laid" is used for designation  of the
meadows at marine coasts; lakes in Vasugany, Siberia; and wet depressions in the tundra.

      These wet areas may be attributed both to terrestrial and aquatic ecosystems which
are diverse and specific.  From the  biogeochemical point of view these areas, when
compared to deep water habitats, are characterized by lower and more variable spacial pH
values, lower concentrations of dissolved oxygen, extreme reducing conditions, a higher
probability of photodegradation and biodegradation, chelation and formation of organic and
organomineral complexes, and higher sulfide concentrations. Hydrology is the controlling
factor in all of these wet areas, in spite of their diversity.  All these ecosystems are
periodically flooded or saturated with water.  In accordance with Polynov classification, these
habitats are attributed  to "aqual" and "superaqual" elementary landscapes (biogenosis).
1Hydrochemical Institute, Federal Survey of Russia for Hydrometeorology and Environmental
 Monitoring, Rostov-on-Don, Russia.
2U.S. Environmental Protection Agency, Environmental Research Laboratory, Duluth, MN.
                                        175

-------
Wetland, which is a new term in Russian scientific literature, describes a diversity of zones
lying between typical terrestrial habitats and deep water ones.

       The U.S. Fish and Wildlife Service (USFWS) developed and published the
classification of U.S. wetlands entitled "Classification of Wetlands and Deepwater Habitats of
the United States" (Cowardin et al. 1979). This classification is not the only one in use in the
U.S., but its ecological comprehensiveness and simplicity has provided an excellent
classification tool. Because of this, we believe it would have an excellent possibility for
classifying Russian wetlands.

       The USFWS system standardizes wetland classification and provides the basis for a
wetland inventory.  There is no unified classification of wet habitats in Russia.  The first
project undertaken by American and Russian specialists was to evaluate the USFWS
classification in  Russia and determine its applicability to Russian wetlands.  These
investigations were carried out in the framework of Cooperative American-Russian  Project
02.02.14. "Influence of Pollutants on Wetland Ecosystems" in accordance with an
Intergovernmental Agreement in the Field of Environmental Protection.

       Below is a brief review of the application of the USFWS wetland classification system
used in Russia.  This is preliminary and the work will be continued.  Some details  of USFWS
classification are omitted purposefully to underline the  principles of classification.
          BRIEF ANALYSIS OF THE USFWS CLASSIFICATION OF WETLANDS

       Wetland is understood as a wet habitat periodically flooded or saturated with water.
The USFWS defines a wetland as transitional between terrestrial and deepwater ecosystems
with the water table at the surface or close to the ground surface.  To classify an
environment as a wetland it must possess at least one of the three following features: 1) the
area should be  inhabited  by hydrophytes, at least periodically; 2) the surface substrate is not
continuously drained, and hydromorphic soils are formed; and 3) soils are saturated with
water or flooded during some part of the yearly growing season, i.e., the hydrological regime
may be variable-from permanent flooding or saturation to occasional water at the land
surface.

       According to USFWS classification of hydrophytic plants, hydrophytes are plants
inhabiting environments characterized by recurring  anaerobic conditions or close to anaerobic
conditions because of high moisture content.  The substrate of wetlands may not necessarily
be a hydromorphic soil, e.g., sand and gravel beaches, but at least for some period during
the vegetation growing season the substrate must be saturated or flooded.

       Because of the absence of a distinct interface between  dry and wet habitats, there
may not be indisputable determinations of wetland  boundaries. We have to use the term
wetland even though it is  somewhat artificial from a scientific point of view. In  Russian
scientific literature, much  less definite terms are used to describe these areas, i.e., wet boggy
areas, which cannot be taken as synonymous with  the term wetland.
                                        176

-------
       Only definite biogenoses may be attributed to wetlands. This is illustrated by Table 1.
However, the basic features of wetlands differ in different climatic zones of Russia.  In all
climatic zones there are plants and animals that cannot survive saturated or flooded soils.

       Wetland spatial boundaries also differ. If marine, delta, riverine, and  lacustrine
wetlands along moisture  gradients (from deepwater to terrestrial  habitats) are considered,
wetland boundaries will be determined by minimum and maximum water levels during the
vegetation growing season.  One  of the conventional boundaries of riverine and lacustrine
wetlands is the distribution of emergent vascular plants (for lacustrine wetlands this is the
2 m depth zone). Beyond this is the deepwater habitat typical of aquatic ecosystems. The
boundaries between terrestrial ecosystems and  wetlands are the areas in which hydrophytes
are replaced by mesophytes and xerophytes. There are also signs of variation in
hydrological regime indicating occasional flooding or soil saturation.

       Table 2 presents USFWS classification with the hierarchical structure of wetlands at
three groups: systems, subsystems, and classes.  Experience with the USFWS classification
system in Russia (Caucasus, southern part of the country; Rostov region, forest steppe zone
of Voronezh; Kursk regions,  taiga of the European part of the country; and basins of the
Olekma and Amga rivers in Yakutia, Siberia) indicated that at the levels of systems,
subsystems,  and  classes,  the system can be applied for classification of Russian wetlands.
No  wetlands  endemic to Russia could not be classified at the level of class or higher. Some
Russian wetlands did  not appear to meet the classification criteria  at the level of subclass, so
it may be necessary to develop specific subclasses for Russian wetlands.
  TABLE 1.  BASIC FEATURES OF WETLANDS"
  Lacustrine
  Palustrine
                        With hydrophytes
Phragmites
Scirpus, Carex,
Salix and others

Carex, Salix,
Phragmites,
Scirpus and
others

Peat bogs
                                             Without hydrophytes
Wetland
system
Marine
Delta
Riverine
With hydric
soils
Rocky shore

Aquatic bed
Without
hydric soils

Plavni
Phragmites,
With hydric
soils
Sand beaches

Temporal
Without
hydric soils
Salt marsh


                                                       streams
Man-made
pool banks
'Absence of wetland terms in some columns does not mean there are no wetlands with such
features.
                                         177

-------
TABLE 2. CLASSIFICATION OF RUSSIAN WETLANDS AT LARGE TAXONOMIC GROUP LEVELS
Systems
Marine

Delta

Subsystems
Subtidal
Tidal
Subtidal
Tidal
Classes
Rock bottom; unconsolidated bottom; aquatic bed
Aquatic bed; rocky shore; unconsolidated shore
Rock bottom; unconsolidated bottom; aquatic bed
Aquatic bed; stream bed; unconsolidated shore;
emergent wetland; shrub wetland; forested wetland
  Riverine       Tidal



               Lower perennial


               Upper perennial


               Intermittent

  Lacustrine     Limnetic

               Littoral



  Palustrine
Rocky bottom; unconsolidated bottom; aquatic bed;
rocky shore; unconsolidated shore; emergent
wetland

Rock bottom; unconsolidated  bottom; aquatic bed;
rocky shore; emergent wetland

Rock bottom; unconsolidated  bottom; rocky shore;
unconsolidated shore

Stream bed

Rock bottom; unconsolidated  bottom; aquatic bed

Rock bottom; unconsolidated  bottom; aquatic bed;
rocky shore; unconsolidated shore; emergent
wetland

Rock bottom; unconsolidated  bottom; aquatic bed;
unconsolidated shore; moss-lichen wetland;
emergent wetland; shrub wetland; forested wetland
       The structure of the higher levels of the USFWS classification system, presented in
Table 2, are unified and may be used for description of different wetlands.  This could not be
achieved using common geographic and geochemical systems for classification of
landscapes.

       Table 2 also shows that the highest taxonomic level in the USFWS classification
scheme is system.  This is determined as  a complex of wetlands exposed to the influence of
similar hydrological, geomorphological, and biogeochemical factors.  According to the
USFWS classification there are five wetland types at the system level:  marine, delta, riverine,
lacustrine, and palustrine.  These systems, with the exception of palustrine,  are subdivided
into subsystems  based on hydrological and geomorphological factors.  As a result, marine
and delta systems possess two similar subsystems: 1) located below the level of the tidal
zone and those flooded  permanently and 2) located above the level of the tidal zone and
occasionally flooded and exposed.
                                         178

-------
       The riverine system is subdivided into four subsystems.  These are enumerated from
the mouth to the head of the river: 1) tidal, exposed to the influence of tides and wind
induced tides; 2) lower perennial with features typical of plains rivers; 3) upper perennial with
features typical of mountain rivers; and 4) intermittent, flooded occasionally during a year.

       The Lacustrine system also comprises two subsystems:  1) limnetic and 2) littoral
subsystem.

       Subsystems of wetlands are divided into classes based on life forms and plant cover.
While identifying classes of wetlands, two additional factors should be taken into
consideration: 1) structure of the substrate cover if there are no soils and 2) water chemical
composition.  These factors will largely determine vegetative composition.  In case plant
cover is less  than 30% of the wetland area, this wetland may be included in one of the
following five classes: 1) rock bottom, 2) rocky shore, 3) unconsolidated bottom, 4)
unconsolidated shore, and 5) stream bed.

       Where plant cover occupies greater than 30% of the wetland area,  the wetland may
be included in the following five classes: 1) aquatic bed, 2)  moss-lichen, 3) emergent
vascular plants, 4) shrub, and 5) forested.

       These classes are subdivided into subclasses (Table 3) identified by detailed analysis
of the life forms of plants and a sublayer of the soil.  The following subclasses were  isolated
in Russia, based primarily on the plant character  and life forms of dominant plants: 1)
persistent, 2) nonpersistent, 3) emergent, 4) moss, 5) lichen, 6) broad-leaved deciduous, 7)
broad-leaved evergreen, 8) needle-leaved evergreen, 9) deciduous, and 10) dead.

       On the basis of the substrate sublayer, classes are divided into the following
subclasses: 1) bedrock, 2) rubble, 3) cobble-gravel, 4) mud, 5) sand, and 6) organic matter
(detritus).

       Subclasses are further subdivided into dominance types, determined by dominant
plant and animal species.  The dominance types  describe individual features of each wetland
and the number of dominant types is indefinite.

       Four types of modifiers are used to more accurately identify characteristics of
wetlands: 1) water regime, 2) water chemical composition, 3) soil, and 4) special.

       Water regime modifiers describe flood conditions or  the degree of wetland saturation.
In this case they are divided into two basic groups: 1) subtidal, intermittently exposed,
permanently flooded, and intermittently flooded; and  2) permanently flooded, intermittently
exposed, semipermanently flooded, seasonally flooded, and artificially flooded.

       Some delta and riverine wetlands are exposed to the influence of tidal and  nontidal
hydrologic regimes. It is difficult to determine flooding patterns and saturation levels for such
wetlands.

       Water chemical composition modifiers consist of two parameters: mineralization  and
pH. These are divided into two groups: 1) marine and delta wetlands exposed to tidal and
                                         179

-------
wind induced influence and 2) riverine, lacustrine, and palustrine wetlands not exposed to
tidal and wind induced influence.
       Soil modifiers are developed for two groups of soils: those rich in organic matter
(more than 20% in a 40 cm layer) and mineral soils.
       The following specific modifiers describe anthropogenic and zoogenic impact on
wetlands: excavated, impounded, ditched, partially drained, farmed, and artificial.
TABLE 3. CLASSES AND CORRESPONDING SUBCLASSES OF RUSSIAN WETLANDS.
   Number
Classes
Subclasses
      1       Rock bottom
      2       Unconsolidated bottom
      3       Aquatic bed

      4       Streambed

      5       Rocky shore
      6       Unconsolidated shore
      7       Moss-lichen
      8       Emergent vascular plants
      9       Shrub

     10      Forested
                    Bedrock, rubble
                    Cobble-gravel, mud, sand, organic matter
                    Algal, aquatic moss, rooted vascular,
                    floating vascular
                    Bedrock, rubble, cobble-gravel, sand, mud,
                    organic matter
                    Bedrock, rubble
                    Cobble-gravel, moss-lichen
                    Moss, lichen, moss-lichen
                    Persistent, nonpersistent
                    Broad-leaved deciduous, broad-leaved
                    evergreen, needle-leaved evergreen, dead
                    Broad-leaved deciduous, broad-leaved
                    evergreen, dead
                  WETLANDS OF THE AMGA RIVER BASIN, YAKUTIA
APPLICATION OF THE USFWS CLASSIFICATION SYSTEM
       To compare characteristics of wetland habitats using the USFWS wetland
classification scheme and the physico-geographic description common in Russia, we chose
the Amga River Valley. This river flows through the Olekma Preserve, Yakutia.
       In July 1991 a 30-day field survey was carried out by the Laboratory of Ecological
Investigations to study the wetlands of this area.
                                        180

-------
       Within the Olekma Preserve, the Amga River has two distinct sections. The first
extends from the headwaters to the inflow of the river Olobochor.  In this section the Amga
River consists of many branches.  There is also a chain of islands along the river. Along the
shore are sand-rubble ridges, and the bed consists  of a shallow pebble bottom.  The second
section extends from the point of the Olobochor River inflow to the boundary of Olekma
Preserve. During this transit, the Amga River Valley becomes wider and the river meanders
extensively.  Many of the meanders are eventually cut off from the river and form bogs within
the floodplain. The area located in front of the wide floodplain terrace contains many bogs.

       Ecological conditions of the Amga River section  correspond to the criteria of the upper
perennial subsystem of the USFWS classification. In the lower section, there are some
features  of the low perennial subsystem: flow current velocity in the pools decrease and  the
proportion of fine fractions in sediment sublayer increases.  In small bays, with the local
name "kuria," the sediment sublayer is mud with organic matter.  Some river tributaries
flooded in the period of peak flow may be classified  intermittent.

       Elements of the Amga River such as stream  bottom, banks, and pools may be
classified as riverine wetlands of the upper perennial subsystem:  class = unconsolidated
bottom; subclass = cobble-gravel (sometimes subclass  = sand or mud); water regime is
permanently flooded, fresh water.

       Elements of the Amga River floodplain composition such as bichevnik (local Russian
term), that is pebble beach, levees, spits, beaches,  summits, and tails of the islands can be
attributed to riverine wetlands of the upper perennial subsystem: class = unconsolidated
shore; subclass = cobble-gravel (sometimes, subclass = sand or mud); water regime  is
seasonally flooded, temporarily flooded, and intermittently flooded, freshwater.

       The section of the Hatyn River at the point of inflow to the Amga River possesses
features of the lower perennial subsystem.  The wetlands at the mouth of the Hatyn River are
class = unconsolidated bottom; subclass = mud.

       Palustrine wetlands were found at the floodplain of the Olobochor River. They were
classified as wetlands of class = shrubs; subclass =  broad-leaved deciduous; dominant genus
= willows (Salix); water regime  is semipermanently flooded, fresh waters. Upstream of the
Olobochor River class = moss-lichen wetlands; subclass =  mosses, possessing saturated
water regime.

       In the vicinity of the point Hatyn, there is a hillocky bog that was classified as class =
emergent; subclass = persistent. In all cases the wetland water was fresh.

       Wet habitats were not limited to the  riverbed  or floodplain of the Amga River.  In the
historical  records of Olekma Preserve, other wet areas  are mentioned:  stream mouths and
willow thickets in the valleys, intermittently exposed  beds of the streams, needle-leaved
deciduous and broad-leaved deciduous forests located  at sedge sphagnum palustrine
wetlands, shrub wetlands, spring wetlands,  etc.  This list does not include all wet habitats of
the area.
                                        181

-------
      Thus, the USFWS Classification tested at the Amga River, Yakutia, allowed us to
place the wet habitats in a hierarchical structure and to describe them with unified terms.
Though the definition of wetlands is sometimes rather long, USFWS classification is still
convenient and may be easily used in Russia if you understand the principles of its structure.
This classification does not need any transformation above the level of subclass.  It may,
however, be necessary to include additional subclasses describing wetlands endemic for
Russia (Euro-Asia).
                                   REFERENCES

Cowardin, L. M., V. Carter, F. C. Golet and E. T. LaRoe. 1979. Classification of wetland and
      deepwater habitats of the United States.  FWS OBS-79/31, Washington, DC.
                                        182

-------
       ECOLOGICAL STATE OF THE UPPER VOLGA RIVER AS INFLUENCED BY
                       COMPLEX ANTHROPOGENIC FACTORS

                                         by

                            N. P. Smimov and A. I. Kopylov1


                                    ABSTRACT

       Ever increasing degradation of wetland and aquatic habitat has caused great urgency
to maintain biological diversity and protect the gene pools of plant and animal populations
native to the Volga River.  Observations point to a sharp decrease in biological diversity and
loss of species-for instance,  many species of aquatic and littoral  plants are on the verge of
extinction.  There is need to identify the degree of anthropogenic  disruption of
structural-functional components of the ecosystem, its ability to restore itself to the original
condition, and potential methods for remediation.


                                   INTRODUCTION

       The Volga River-3530 km long-is the largest river in Europe  and the sixth largest in
the world.  In  general, the Volga is divided into three parts: Upper, Middle,  and Lower.  The
dam of Rybinskoye Reservoir, near the town of Rybinsk, is considered the  boundary between
the Upper and Middle Volga.  The Volzhskaya Hydro-electro Power Station and dam
separates the Middle from the Lower part of the Volga River.

       The Upper Volga, about 800 km long, includes the Verhnevolzhskoye,  Ivankovskoye,
Uglichskoye, and Rybinskoye reservoirs. All the reservoirs,  except the Verhnevolzhskoye,
are about 50 years old and differ significantly in morphometric characteristics (Table 1).

       Today, flora and fauna of the Upper Volga are relatively well studied, due mainly to
investigations of the Institute of Biology of Inland Waters (IBIW), Russian Academy of
Science (Table 2).  There are 1018 species  and varieties of algae and 342 species of higher
aquatic plants. The total number of invertebrate and fish species  found in  reservoirs of the
Upper Volga is 2288, including 62 species that are considered rare.  Insects, nematodes, and
Institute for the Biology of Inland Waters, Russian Academy of Science, Borok, Nekouz,
 Yaroslaval, Russia.
                                       183

-------
TABLE 1.  PRIMARY MORPHOMETRIC CHARACTERISTICS OF RESERVOIRS ON THE UPPER
VOLGA RIVER
Depth
Reservoir
Upper Volga
Ivankovo
Uglich
Rybinsk
Volume
(km3)
794
1,120
1,245
25,420
Area
(km3)
179
327
249
4,550
Length
(km)
92
120
143
250
Mean
(m)
4.4
3.4
5.0
5.6
Maximum
(m)
16.1
19.0
23.2
30.4
Dates of creation
and reconstruction
1843 (1943-47)
1937
1940
1941-47



TABLE 2. FLORA AND FAUNA OF THE UPPER VOLGA RIVER
General groupings
Algae
Vascular plants
Protozoa
Spongia
Coelenterata
Platyhelminthes
Nemathelminthes
Acanthocephala
Annelida
Mollusca
Tentaculata
Arthropods
Chordata
Species numbers
1018
342
139
2
1
240
610
7
79
64
9
720
32
rotifers are the most rich in species. In total, there are more than 3000 species of
autotrophic and heterotrophic eukaryotic organisms in the Upper Volga basin.

       Regular observations of floral and faunal species composition over many years can be
a sensitive indicator, or monitoring tool, to evaluate the state of the ecosystem.  Available
data suggest a sharp decrease in biological diversity  and an irreparable loss of species and
even biocoenosis in the region. Also, the role of invading species from northern waterbodies
seems to be decreasing.  Siltation of flooded channels resulted in the disappearance of
                                        184

-------
mysids and gammarids from some areas.  Changes in hydrophilic flora have been observed,
and many species of aquatic and littoral plants are on the verge of extinction.

       At the same time, unusual species or those previously exotic to the region are
appearing in the reservoirs, or entire drainage of the Upper Volga.  For example, the Caspian
polychaeta (Hypania invalida), common in  southern reservoirs, was found in  Ivankovskoye
Reservoir in 1986.  The ostracod (Stenocirpis  malcolmsoni)~a typical inhabitant of European
botanical gardens and heated basins-has  also been discovered in the same reservoir.  It is
interesting  that uncommon parasites are appearing  in the area. Mass infections of cyprinid
fry  (roach,  Rutilus rutilus; and white bream, Blicca bjoerkna) by the ectoparasite infusorium
(Ambiphrya ameiruri) has been noted in the Rybinskoye Reservoir. This infusorium was
brought to  Europe in the 1980s along with  its host-trie American catfish (Ictalurus
punctatus)~and passed on to other native  species in fish farms of the country.  Invasion of
the ectoparasites into the Upper Volga is apparently linked to conditions favorable for
reproduction of the  infusorium (high concentrations of dissolved organic matter and abundant
plankton microflora), which occur in some areas.  These are the primary foods for this
infusorium.  The ectoparasites are distributed evenly on the skin epithelium and sometimes
cover 60% of the surface.

       Significant changes in the species composition of plankton and benthic communities
are primarily the result of eutrophication and pollution from anthropogenic sources in the
Upper Volga basin.  Anthropogenic loads on waterbodies have greatly increased within the
last decade.  Primary anthropogenic sources include industrial and municipal wastes released
into waterbodies, air pollution, more intensive use of the Volga basin for transportation and
recreation purposes, industrial construction along the banks of the river and it's basin,
dredging for construction purposes,  and intensive use of biological resources.

       This manuscript presents just a small part of the information on the effects of
anthropogenic eutrophication and anthropogenic pollution of the Upper Volga reservoirs
gained by scientists of the IBIW.
                   EUTROPHICATION OF THE UPPER VOLGA RIVER

       Reservoirs are unique aquatic ecosystems where biological succession is a
complicated and infinitely long process. The succession, or aging, of reservoir ecosystems
differs markedly from that of lakes. Anthropogenic factors  affect reservoirs from the moment
of their completion and play a lead role in reservoir ecosystem dynamics.

       Trophic level changes in different stages of reservoir succession are caused by a
variety of factors and are not always associated with anthropogenic enrichment by biogenes.
Therefore, eutrophication of reservoirs is not always caused by anthropogenic factors. Thus,
it is more correct to consider changes in the structural and functional characteristics of
aquatic communities to be the result of a combination  of natural and anthropomorphic
eutrophication.  Ivankovskoye Reservoir is a good example to illustrate this combined
eutrophication process.  In addition to diatoms, pyrrophytic and blue-green algae begin to
dominate in  phytoplankton communities, and can be used as indicators of eutrophic
conditions.  Before 1970, the blue-green algae biomass had not exceeded 7 g/m3, but within
                                        185

-------
the last decade it has reached hundreds of grams per cubic meter.  Intensive increases in
these algal species groups have become characteristic of all reservoirs in the region. In
comparison to the 1950s, total biomass of plankton has increased by 2.0-6.6 times.  Data on
phytoplankton primary production from the  1980s shows that production in the photosynthetic
zone has increased by 1.2-6.5 times during the same period.

       Along  with phytoplankton, total numbers and biomass of zooplankton increased
significantly over the last few years. The number of species that are indicative of
eutrophication has increased sharply as  has filter feeding zooplankton. Also, ratios between
taxonomic groups have changed.

       The present state of eutrophication  in Ivankovskoye Reservoir is characterized by
increases in annual macrophyte production, and percent area of coverage by emergent
plants. Simultaneously, the process of endogenous succession is resulting in increased
dominance of plants typical of bogs.  The appearance of herbaceous floating mats followed
by replacement with brushy and woody plants is common for this type of endogenous
succession. The following scheme illustrates the change in plant communities in general:
1) emergent plants; 2) emergent plants and floating mats; 3) floating mats and islands; and,
finally,  4) floating islands with individual willow shrubs and trees, and birch and fir forest.  At
the willow development stage, shallow waters are converted into boggy soil, over a period of
30-40 years.  The Ivankovskoye  Reservoir  has been  in existence for 40 years, and due to
endogenic succession, floating mats now cover 14%  (13 km2) of littoral area, and 3%
(11  km3) of the total area of the reservoir has been turned into boggy soil (Table 3).
TABLE 3. HIGHER AQUATIC PLANTS OF IVANKOVSKOYE RESERVOIR


Observation                    1957                1973               1980


Area of vegetation (km2)           55                   79                 83

Degree of overgrowth (%)         17                   24                 28

Annual production of organic
 matter (higher water plants,
 thousands of tons)               25                   54                  68
                  POLLUTION OF THE UPPER VOLGA RESERVOIRS

       Effects of discharge of large amounts of industrial and municipal wastewater have
been studied in the northern part of the Sheksna Reach of the Rybinskoye Reservoir (near
the town of Cherepovets).  Major pollutants include oil products, polychlorinated biphenyls
(PCBs), di- and polyaromatic hydrocarbons (PAHs), heavy metals, and others. According to
1990 data, taken from some locations, the concentration of oil exceeded the maximum
                                        186

-------
permissible concentration by four times, and the concentration of PAHs by 150 times (Table
4).

       The pollution of bottom sediments by PAHs presents a serious hazard. The
concentrations, extrapolated from benzopyrene, for the majority of samples were 25-53 mg/g
of dry weight, which exceeds by 3-4 orders of magnitude the background concentration in
unpolluted areas.  Bottom sediments and fish  also contained a number of genotoxic
substances which are known to cause mutagenic effects. The portions of Skeksna Reach
receiving these wastes should be considered toxic to aquatic resources.  During summer, the
length of this pollution zone is about 7 km.  As oxidation/reduction transformations of the
various pollutants occur at this time, the toxic environment expands into a zone of intensified
eutrophication  that extends for 20-25 km downstream.

       Ecological  impacts of pollutants can be observed at the organism, population, and
community levels.  The disturbance of some physiological functions, changes in behavior,
reduced growth rates, and increased morbidity and mortality are caused by direct poisoning
and decreased tolerance to different stresses.  Aquatic organisms in toxic zones  accumulate
these pollutants (Table 5).  According to 1990 data, the concentrations of persistent and
accumulable PCBs were higher in invertebrates than in bottom sediments.  Fish containing
PCB residues were found in areas of the reservoir considerable distances away from the
zone directly polluted by this substance. In addition, we have observed a number of
moribund mollusks (Unio sp.) with different stages of liver degeneration  and glandular cancer.

       Fish captured in the polluted sections of Skeksna  Reach were found to have a
60-90% incidence of immune system deficiency, whereas fish taken from habitats remote
from Cherepovets had an incidence of 0-30%.  The livers and tissues associated with the
immune system of immuno-deficient fish showed severe cirrhosis and had little natural
immunity to disease.

      Crayfish (Bythotrephes longimanus) collected in the reach were moribund, and
examination showed severe degeneration of the intestinal epithelium and abnormal
TABLE 4. POLLUTANTS IDENTIFIED IN WATER AND BOTTOM SEDIMENTS IN THE NORTHERN
PART OF THE SHEKSNA REACH (RYBINSK RESERVOIR, NEAR CHEREPOVETS 1990-91)

                                        Water                    Bottom sediments
Pollutants                                (//g/L)                    (mg/g dry weight)
Polychlorinated biphenyls                  Trace = 0.29                Trace = 0.60

Oil products:
  Hydrocarbons                             0.05-0.19                 0.56-0.62
  Resinous components                      0.00-0.03                 0.44- 2.65

Polyaromatic hydrocarbons                    0.10-0.64                 3.33-49.73
                                       187

-------
development of the placental structure leading to the brood pouch.  This can lead to mortality
of parthenogenetic embryos developing in the pouch.  As a result, crayfish reproduction and
population levels have declined, and its distribution has narrowed. In the near future, we
expect crayfish to disappear altogether.

       At the population level, pollution causes changes in number and biomass, birth and
mortality, and in dimensional structure and dynamics in aquatic organisms. The
concentrations of saprophytic bacteria, as well as specialized groups of microorganisms
which oxidize phenol, naphthalene, oil, and other organic pollutants, increases sharply in
polluted waters.  The presence of these toxicants in water has an especially negative impact
on large crayfish with prolonged life cycles (Bythotrephes sp., Heteroscope sp., Eudiaptomos
sp.). Genetic and population-level analyses have shown considerable disturbance in the
genetic structure of fish inhabiting areas receiving industrial wastewaters.  These genetic
changes are similar to those observed in the zones of radioactive pollution at the Chernobyl
Nuclear Station.
TABLE 5. RESIDUES OF POLLUTANTS IN TISSUES OF FISH AND BENTHIC INVERTEBRATES IN
THE NORTHERN PART OF THE SHEKSNA REACH (RYBINSKOYE RESERVOIR, NEAR
CHEREPOVETS, 1990-91)
Organism
Tissues
   PCS*
  (mg/kg)
PAC + OPCT
 (mg/kg)
Oil products
  (mg/kg)
White bream
Pike perch
 (Lucioperca
 vo/gens/s)

Blue bream
 (Abramis ballerus)
Dreissena

Chironomidae

Oligochaetes

Viviparus
 (pulmonate snails)
Muscle
Liver

Muscle
Liver
Muscle
Liver
 FISH

0.00-0.53
0.45-5.41

0.00-0.49
0.77-1.76
0.25
0.87
                INVERTEBRATES

Whole organism       0.20-1.40

Whole organism       1.15-0.67

Whole organism       0.00-0.42

Whole organism       0.14-1.32
3.60-39.4
0.40-23.3

15.4
20.0
  14.465
  10-140

  17.0
  17.0
                 9.0-37.0
                  39.0-733.0
*PCB = polychlorinated biphenyl.
tPAC = Polyaromatic hydrocarbon; OPC = oxidized polynuclear compound.
                                        188

-------
       At the ecosystem level, pollution affects the structure and function of plankton and
benthic communities.  Areas, such as the Sheksna Reach, receiving heavy loads of industrial
and municipal wastes from Cherepovets, characteristically have a low index of species
diversity and a low quantity and biomass of phytoplankton and zooplankton (Table 6).
Trophic food chains are shortened in these aquatic communities and the number of predators
has declined sharply.  In spite of extremely high concentrations of biogenic elements, the
intensity of photosynthesis in these areas is low, which is probably related to low water
transparency and toxic effects of wastewaters.  The significant increase of bacterial
production  over primary production is the likely result of the intensive destruction of microflora
and microfauna that feed on bacteria, such as rotifers, etc.  Such increases in development
of planktonic bacteria  in comparison to phytoplankton is related to the input of large amounts
of anthropogenic organic matter.  These relationships are demonstrated  by ratios of primary
production  to total heterotrophic decay (Table 6). South of the Cherepovets area, toxic
effects of pollution decrease and eutrophication occurs: photosynthesis increases sharply and
total biomass of the plankton community increases.  An increase in species diversity is also
observed, but the role of predators in the total biomass of zooplankton remains insignificant.

       Pollution of bottom sediments greatly influences the structure and function of benthic
communities.  Monitoring  of microbiological processes taking place in the bottom sediments
TABLE 6. STRUCTURAL/FUNCTIONAL CHARACTERISTICS OF PLANKTON COMMUNITIES IN THE
POLLUTED SHEKSNA REACH AND TWO DOWNSTREAM ZONES OF THE RYBINSKOYE
RESERVOIR, SUMMER 1989

                                                               Downstream zones
Parameters
Zooplankton
(Shannon Index) t
Total plankton biomass
Ratio, biomass of predator
zooplankton to prey (%)
Sheksna Reach*
1.4-1.7
7.1
22.2
Eutrophic
3.9
17.0
14.0
Relatively clean
2.0-2.5
3.8
84.7
Photosynthesis
  (mg carbon [C]/M3/hour)            0.94                      2.21          0.77

Bacterial production                 2.97                      2.10          1.20
  (mg C/m3/hour)

Ratio, primary production to
  total heterotrophic decay           0.45-0.73                 0.44-1.51     2.24-3.73


*Data from 1986-88.
t Index of biodiversity.
                                        189

-------
of the toxic zone can be used as an indicator of the adverse effects of long-term chronic
pollution, such as that caused by wastes discharged in the Cherepovets area.  The specific
activity of dark assimilation of C14 tagged carbon dioxide in sediments collected in the
polluted zone was higher than in sediments from other regions of the reservoir (Table 7).
These results suggest that certain microflora in sediments function actively to break down the
different pollutants in spite of their low abundance.  In contrast to less polluted regions, there
is a sharp decrease in aerobic processes and  an increase in anaerobic methane-forming and
sulfate-reducing microflora.  Thus, the pollution-induced destruction of the aerobic bacterial
community, and development of a stable anaerobic community has seriously inhibited natural
self-purification processes in the sediments, and further, has caused secondary pollution in
the form of anaerobic decay products in downstream waters.

      Anthropogenically caused changes in the Sheksna Reach of the Volga have been
caused by heavy  industrial discharges, and the situation can be considered an ecological
disaster.  It is obvious that the return of this ecosystem to its original condition through
self-renewal is impossible without compulsory  remediation.  The development of local zones
with varying degrees of adverse changes in their structural-functional ecosystem components
can be found near other large industrial centers. In addition, the riverine ecosystems
downstream of these polluted zones can be considered "at ecological risk" due to increased
eutrophication.  However, it is possible to return these ecosystems to their original condition
naturally by remediating upstream pollution.
TABLE 7. FUNCTIONAL CHARACTERISTICS OF BENTHIC MICROFLORA IN THE POLLUTED
SHEKSNA REACH AND TWO DOWNSTREAM ZONES IN THE RYBINSKOYE RESERVOIR, SUMMER
1986-89

                                                             Downstream zones
Parameters                      Sheksna Reach            Eutrophic    Relatively clean


Dark assimilation                    2.7                     2.0              1.8

Specific activity per 1 x 10s
  celts of bacteria                   1.7                     0.5              0.2

Aerobic decay,
 mgC/rr»2/day                      87                     203             18O

Methane transformation,
  mt/dm3/day

       Formation                   2.5                     0.06
       Oxidation                   1.6                     0.01

Intensity of sulfate
  reduction, mg S/kg/day           41.1                     3.1              1.1
                                         190

-------
                                   CONCLUSIONS

       With a growing trend of wetland and aquatic habitat degradation, there is even greater
urgency to maintain the biological diversity and protection of gene pools of aquatic plant and
animal populations native to the Volga River.  The present situation requires enhancement of
research to estimate the tolerance (resistance)  of biological systems and frequency of their
anthropogenic-caused changes.  Complex ecological investigations of areas under the
influence of industrial-municipal wastes have special significance because they permit us to
identify the degree of anthropogenic disruption  of structural-functional components of the
ecosystem, its ability to restore itself to the original condition, and to evaluate potential
methods of remediation.
                                        191

-------
                     RESTORING AQUATIC RESOURCES TO THE LOWER
                          MISSOURI RIVER: ISSUES AND INITIATIVES
                                              by

                     David L. Galat1, John W. Robinson2 and Larry W. Hesse3
                                         ABSTRACT

       Large, low-gradient rivers, such as the Volga and the lower Mississippi, are dependent on the
river-floodplain linkage and periodic flood pulse for their integrity. The Missouri River is the longest
river in the United States and in presettlement times exhibited extensive channel migration and high
turbidity.  The flora and fauna of the river-floodplain complex reflected this physical dynamic equilibrium.
Presently, about one-third of the river has been channelized, one-third impounded, and the  hydrograph
of the remaining one-third severely altered.  Additionally, agricultural,  industrial, and urban development
in the river basin has degraded water quality.  Physical and chemical consequences of these alterations
are reviewed as well as impacts on biological structure and function, including commercial and sport
fishery resources.  Restoration of the Missouri River necessitates a return of the ecosystem to an
approximation of its former condition, including recreating both structural and functional components.
Objectives to attain this goal include reestablishment of a semblance of the precontrol hydrograph and
sediment regime, restoration of some of the precontrol channel structural diversity and the river-
floodplain linkage, reestablishment and enhancement of native Missouri River fishes and their
migrations, reduction of major point and nonpoint sources of pollution, and restoration of native
terrestrial and wetland plant communities along the river channel and floodplain. Several recent
initiatives are summarized that attempt to resolve competing watershed interests and address one or
more of these objectives.  These initiatives are the Mississippi Interstate Cooperative Resource
Agreement, the Cooperative Interjurisdictional Rivers Fisheries Resources Act of 1992, the Missouri
River Initiative (MOR-CARE), and the Missouri River Fish and Wildlife Migration Project.  Successful
restoration of the Missouri or any river-riparian ecosystem necessitates a holistic perspective at the
landscape scale of the river basin.
1Missouri Cooperative Fish and Wildlife Research Unit, U.S. Fish and Wildlife    Service, 112
Stephens Hall, University of Missouri, Columbia, MO 65211,     USA

2Missouri Department of Conservation, Columbia, MO 65201, USA

3Nebraska Game and Parks Commission, Norfolk, NE 68701, USA
                                             192

-------
                                   INTRODUCTION
       Most large rivers in developed countries have been severely influenced by human
alteration (Petts 1984, Davies and Walker 1986, Hynes 1989) and the Missouri River is no
exception.  Significant Anglo intervention began after the Louisiana Purchase, when in 1804
Lewis and Clark were commissioned by the federal government to find a road to the west for
economic development (Keenlyne 1988). Subsequently, the Missouri River became the first
great highway for exploitation and settlement of the American west.  The Missouri River today,
like the Volga, has been so radically altered by damming, channelization, and pollution that its
fundamental aquatic character and processes no longer approximate natural conditions.
Our goal is to share with our Russian colleagues existing information on the lower Missouri
River.  To accomplish this goal we have four objectives: 1) review the theoretical framework
for perceiving large river-floodplain ecosystems and natural versus man-induced disturbance,
2) briefly describe the lower Missouri River ecosystem, 3) summarize major alterations to it
and their effects on the biota, and 4) conclude by recommending restoration approaches and
reviewing current restoration efforts.  We believe that despite obvious differences between the
Missouri and Volga rivers they share the fundamental similarities of all large rivers. Moreover,
both rivers have experienced comparable alterations (Hesse et al. 1989a, Pavlov and Ya
Vilenkin 1989).  Collaborative research is essential to achieve our mutual goals of
understanding and restoration of both systems. Exchange of information is the first step to
facilitate collaborative research (Biette et al. 1989) and to develop the knowledge or
approaches needed to restore and manage the aquatic communities of these large rivers.
                  RIVER-FLOODPLAIN INTERACTIONS IN LARGE RIVERS
       Disturbance and recovery of large rivers cannot be understood without a conceptual
framework of their normal behavior. Streams and rivers exist in a state of dynamic equilibrium
(National Research Council 1992).  Local physical features are naturally created, change
through time, and eventually disappear, while the overall pattern (e.g., riffle-pool sequence,
meandering) remains constant at large spatial and long temporal scales.  This dynamic
equilibrium in the physical system creates a corresponding dynamic equilibrium in the
biological system.

       Contemporary perceptions of the structural and functional properties of lotic waters
are largely expressed in two paradigms: the River Continuum Concept (RCC, Vannote et al.
1980, Minshall  et al. 1985) and the resource spiraling concept (Webster and Patten  1979,
Newbold et al.  1981, Elwood et al.  1983). The RCC says that a continuous gradient of
physical conditions and resources  exists from a river's headwaters to its mouth. The stream's
physical features provide much of the  habitat templet for stream community structure and
function. River networks are viewed as longitudinally connected systems  of ordered biotic
                                         193

-------
assemblages, forming a temporal continuum of synchronized species replacements.
Ecosystem level processes in downstream reaches are linked to those upstream through
processing inefficiencies or leakage so that upstream energy loss becomes downstream
energy gain.  Consequently, there is a trade-off between maximizing nutrient and energy use
within a reach via retention mechanisms that minimize downstream  energy loss and the
dependency on this material to drive downstream processes.

       Within this framework a storage-cycle-release phenomenon termed resource spiraling
(rather than recycling) becomes apparent because of the unidirectional flow of water and
continuous transport of materials in lotic ecosystems  (Webster and  Patten 1979, Elwood et al.
1983).  Efficiency of utilization of nutrients and organic carbon within a reach is associated
with the tightness and magnitude of the spirals.  Physical retention, microbial activity, and
macroinvertebrate processing are important activities  for defining the tightness of resource
spiraling and preventing rapid throughput of materials (Minshall et al. 1985).

       The RCC and resource spiraling concepts were developed largely from small
temperate streams, and their usefulness as generalized paradigms for large rivers has been
questioned (Davies and Walker 1986, Statzner and Higler 1985, Welcomme 1985,  Cuffney
1988, Junk et al. 1989, Sedell et al. 1989). A more relevant framework has now emerged in
which to test and  clarify concepts about the structure and function of large floodplain river
ecosystems (Dodge 1989). This complementary perspective is termed the flood pulse
concept (Junk et al. 1989). Junk et al. postulate that  the bulk of aquatic biomass in many
unaltered large floodplain rivers  is derived directly or indirectly from production within the
floodplain and not from downstream transport or organic matter produced elsewhere in the
basin.  Whereas longitudinal linkages in small to moderate sized streams are the basis for  the
continuum  aspect within the RCC, lateral exchange between the floodplain and river channel
and nutrient recycling within the floodplain have a more direct impact on the biota and
biological activity in large  rivers. Whereas downstream losses of organic matter in small
streams are reduced primarily by instream structure (e.g., pools, debris dams), geomorphic
features within the lateral floodplain (e.g., sloughs, side channels, backwaters)  are largely
responsible for retention of organic matter and nutrients in  large low-gradient rivers. The
foundation of the flood pulse concept is that seasonal pulsing of flood flows onto the
floodplain is the driving force controlling the  river-floodplain complex (Junk et al. 1989,
Welcomme et al. 1989, Sparks et al.  1990, Bayley 1991, Schlosser 1991).

       While contributions of organic matter from floodplains may be quantitatively smaller
than from upstream sources, they may be nutritionally of higher quality.  Fremling  et al. (1989)
postulate that organic matter from tributary sources consists largely of dissolved humic acids
or refractory particles by the time it is delivered to the main stem Mississippi River. The more
nutritious fractions have been utilized or retained by upstream communities. They conclude
that local sources of primary production, largely from  within the floodplain,  are  responsible for
the high fish production observed  in large floodplain rivers.

       Floodplain wetlands are regarded as among the most productive ecosystems in the
world (Lieth and Whittaker 1975, Brinson et al. 1981).   In situ  primary production is high, and
effective retention mechanisms contribute to efficient internal  recycling of most carbon and
nutrients (Junk et al. 1989). Although nutrient and organic matter losses from the floodplain
complex to the river channel may be small relative to  internal inputs within the floodplain,
leakage from the floodplain to the river during the annual flood pulse is the principal source of


                                         194

-------
           level of
           floodplain —«*•

          FLOODPLAIN
          SURFACE

        FLOOD;'LAIN
        DEPRESSIONS
                RIVER
  INUNDATED

CONNECTED TO
     RIVER
CONNECTED TO
  FLOODPLAIN
    TERRESTRIAL

ISOLATED EPHEMERAL
 AND PERMANENT
 AQUATIC HABITATS

    CONFINED TO
     CHANNEL
X ^^^^




"7 ^^^^MH
                     8
                        10
                 13
Fig. 1. Idealized changes in water level over an annual cycle for a riverine floodplain.
      Numbered horizontal bars indicate characteristic annual periodicity patterns for some
      major interactions as follows: 1) nutrients released as floodplain surface is flooded; 2)
      nutrient subsidy from river; 3) rapid growth of aquatic plants and invertebrates on
      floodplain; 4) major period of detrital processing on floodplain; 5) DOM and FPOM
      exported to river; 6) maximum plankton production in floodplain depressions; 7) drift of
      plankton, benthos and macrophytes to river; 8) fishes enter floodplain from river and
      fishes that survived dry season in floodplain depressions move to floodplain  surface;
      9) major period offish  spawning on floodplain; 10) period of maximum fish growth; 11)
      fishes move from floodplain to river; 12) heavy fish predation losses at mouth of
      drainage channels; 13) high mortality of fishes stranded in floodplain depressions.
      (Source: Ward 1989.)
these materials to the main channel in unaltered rivers (Mulholland 1981, Cuffney 1988,
Grubaugh and Anderson 1989, Junk et al. 1989, Ward 1989, Sparks et al. 1990).

      Fishes capitalize on this highly productive floodplain environment for feeding,
spawning, nurseries, and as refuge from adverse river conditions (Fig. 1). Indeed, floodplain
wetlands are considered the essential component responsible for the high fish production
recorded in large, low gradient-rivers (Welcomme 1985, Ward 1989).  Risotto and Turner
                                       195

-------
                   Stage

                   use by fish
              Temp.
 0)
 O)
 2
 CO
     Jan.
CL
I
                      
-------
                    NATURAL AND ANTHROPOGENIC DISTURBANCE
       Disturbance in lotic waters has been defined as any unpredictable, discrete event that
disrupts structure or function at the ecosystem, community, or population level (Resh et al.
1988). Lack of predictability is an important component of this definition. Resh et al. (1988)
consider disturbances as events characterized by a frequency (rate of occurrence of events)
and intensity (physical force of event per time) that are outside a predictable range (see Poff
[1992] for an alternative perspective).  Periodic flood pulses of large rivers are predictable
events under this definition.  Indeed, the periodic flood pulse is critical to maintenance of
aquatic populations, communities, and ecosystem processes.  From this perspective, floods
are not disturbances,  unless so amplified, reduced,  or mistimed that they fall outside the long-
term pattern (Sparks et al. 1990).  Large floods are not disruptive events in the long term, as
they contribute to the dynamic equilibrium of the system. Such flood events can  reset late
successional stages to earlier stages, thereby increasing habitat and species diversity (Sparks
et al. 1990). Sand islands are an example of such an ephemeral early successional habitat in
the Missouri River.  Grace (1985) reported 46 species, or two-thirds of the total lower Missouri
River fish fauna, utilized this habitat. Also, two federally listed birds,  the least tern (Sterna
antillarum)  and piping plover (Charadrius melodus) nest primarily on sand islands.

       Man has isolated rivers from their floodplains by draining and filling wetlands,
channelizing river segments, constructing levees to contain flood flows within the main
channel, and constructing mainstream dams and impoundments to reduce downstream
flooding and regulate  flow (Petts 1984, Brookes 1988, Ward and Stanford 1989, Bayley 1b91).
These activities have drastically affected aquatic communities and processes and severed the
river-floodplain linkage.  Channelization and damming, together with agricultural, municipal,
and industrial pollution, constitute the major man-induced disturbances to the integrity of the
world's large river ecosystems.  We will briefly describe the nominal state of the Missouri River
and then summarize effects of these disturbances.
                             MISSOURI RIVER ECOSYSTEM
       The Missouri River's present southeasterly diagonal course across the midcontinent of
the United States (U.S.) traces the southern limits of Pleistocene glaciation (Fig. 3).  It is the
longest river in the U.S., 3,768 km, with a drainage basin encompassing about 1,327
thousand km2 or about one-sixth of the continental U.S. Four physiographic provinces
comprise its drainage basin: 142 thousand km2 of the Rocky Mountains in the west, 932
thousand km2 of the Great Plains in the center of the basin, 228 thousand km2 of Central
Lowlands in the north lower basin, and 24.5 thousand km2 of the Interior Highlands in the
south lower basin (Slizeski et al. 1982, Robison 1986).  River slope varies from about 38 m
km'1 in the Rocky Mountains to an average of 0,17 m km"1  in the Great Plains and Central
Lowlands (USAGE 1985). A prominent feature of the Missouri River's drainage pattern is that
most major tributaries in the upper and middle portions of  the basin  enter on the right bank,
flowing to the east or northeast.

       Climate of the basin is controlled by three air circulation patterns: one originating in
the Gulf of Mexico, another in the northern  Pacific Ocean, and the third in the northern polar


                                          197

-------
Canyon Ferry Reservoir
 (RK 3.688)       .
                                        -Fort Peck (Fort Peck Dam) (RK 2.851)

                                             . Sakakawea (Garrison Dam) (RK 2.237)

                                                       • Oahe (Oahe Dam) (RK 1,725)

                                                          Sharp* (Bif Bend Dam) (RK 1,588)

                                                            Francis Case (Fort Randall Dam) (RK 1.416)


                                                                Yankton(RK 1,295)

                                                             Lewis and dark (Gavins Point Dam)
                                                                      (RK UOS)
                                                                Iowa
Fig. 3. The Missouri River Basin showing most of the civil works projects completed by the
       U.S. Army Corps of Engineers and U.S. Bureau of Reclamation. RK represents river
       kilometer. (Source: Hesse et al. 1989a.)
region (USAGE 1985).  The freeze-free season ranges from fewer than 40 days in the Rocky
Mountains to more than 120 days in the Interior Highlands (Hesse et al. 1989a).  The
drainage basin is generally arid and subject to seasonal and long-term droughts due to the
dominance of the Great Plains physiographic region.  Average annual precipitation ranges
from more than 80 cm  in the Rocky Mountains, to about 45 cm in the Great Plains, and more
than 90 cm in the Interior Highlands (Hesse et al. 1989a).
        Annual Missouri River discharge to the Mississippi River is about 7.0 x 1010 m3.  Two
seasonal periods of flooding occurred prior to impoundment (Fig. 4). The first, or March rise,
was caused by snow melt in the Great Plains and break-up of ice in the main channel and
tributaries.  The second, or June rise, was produced by runoff from snow melt in the Rocky
Mountains and rainfall throughout the basin.

       In presettlement times the Missouri River was one of the most turbid river systems in
North America, earning it the nickname Big Muddy.  The magnitude of the Missouri  River's
sediment load can best be illustrated by its influence on the Mississippi  River.  Average
turbidity of the Mississippi River upstream of the preimpounded Missouri River was reported
                                          198

-------
           r  "%
            6
            o

            55
            O
600

500 •

400 -

300 •

200

100 ^
            ,-J
            w
600 •

500 -

400 -

300 •

200 -

100 -
                                               -^ X
                          /Vf
                             60     120     160    240    300

                                    JULIAN  DATE
                                                360
Fig. 4.  Mean (solid line), maximum (dotted line) and minimum (dashed line) stages of the pre-
       (upper panel) and post-regulated (lower panel) Missouri River at Omaha, Nebraska
       (Rkm 1107). Pre-regulated stages are based on averages of daily stage recordings
       between 1880 and 1899 and post-regulated stages are based on averages between
       1966 and 1985. (Modified from Hesse and Mestl 1992.)
by Plainer (1946) as 300 ppm, whereas below the mouth of the Missouri River the average
increased to 1,800 ppm.

      The Missouri River's course through highly erodible soils resulted in major changes in
channel configuration during flooding. During normal flows the channel was characterized by
continuous bank erosion, a braided shifting configuration, and numerous sand islands and
bars.  Extensive channel migration in the lower river resulted in a floodplain width of 2.4 to
27.4 km, averaging  8.1 km (Hesse et al. 1989a).  For example, about one-third of the
floodplain of the lower Missouri River was reworked by the river between 1879 and 1930
(Schmudde 1963).

      Erosional and depositional characteristics of this  dynamic equilibrium resulted in a
range of serial forest communities in the Missouri River floodplain (Bragg and Tatschl 1977).
Recently deposited  and exposed sandbars are rapidly colonized by willows fSa//x spp.) and
succeeded by cottonwood (Populus deltoides), which dominates the canopy for up to 30
                                        199

-------
Date
         TABLE 1.  SELECTED CHRONOLOGY OF SIGNIFICANT EVENTS
          IN THE HISTORY OF LOWER MISSOURI RIVER DEVELOPMENT
         (PRINCIPAL SOURCES: FUNK AND ROBINSON 1974, FORD 1982,
HESSE 1987, HESSE ETAL 1982, 1989A, BENSON 1988, SCHMULBACH ET AL 1992)

                                   Event
1803          Acquisition of basin to U.S. from France through Louisiana Purchase.

1804-06       Captains M. Lewis and W. Clark expedition of Missouri River from mouth at St Louis,
              Missouri, to origin in Montana

1819          First steamboat travel on Missouri River.

1829          First commercial steamboat barge line: St. Louis to Leavenworth, Kansas-Steamboat
              era begins.

1832          Snag removal authorized under Act of Congress.

1838          2,245 large trees removed from river channel and 1,700 overhanging trees cut from
              bank in 619 km of river upstream from St Louis.

1867-68       Major C. W. Howell Survey and Report on Improvement of Missouri River.

1869          Peak of steamboat era; 47 steamboats deliver about 9,000 mt of cargo to Ft. Benton,
              Montana, 3,540 km upstream from St. Louis.

1881          Lt. Col. C. R. Suter Report detailing  long-range plans for aiding navigation on river.

1884          Missouri River Commission established by Congress to improve navigation of river by
              contracting its width, stabilizing channel location, protecting banks from erosion, and
              snag removal.

1885-1910     Snag removal systematic and intensive; 17,676 snags, 69 drift piles, and 6,073
              overhanging bankline trees removed in 866 km of river in 1901  alone.

1902          Repeal of Act establishing Missouri  River  Commission.  Railroads dominate freight
              traffic-Steamboat era ends.

1902          Congress enacts Reclamation Act of 1902 (PL 57-161) to survey, construct, and
              maintain irrigation works in arid lands of the western United States.  Start of reservoir
              development planning.

1902-12       No maintenance of Commission structures-most wash out.

1910          Increase in typhoid deaths in towns along Missouri River.

1912          Congress authorizes 1.8 m deep, 61 m wide channel from Kansas City, KS to St.
              Louis, Missouri (PL 62-241).
                                       (continued)


                                          200

-------
Table 1. (continued)
Date
Event
1912-17       Active dike and revetment construction to stabilize channel.

1913         U.S. Public Health Service (USPHS) report identifies sewage pollution in river as a
              major factor in typhoid deaths.

1917-33       Maintenance of channel structures, active period of levee construction.

1920-58       Records and studies of water suppliers and USPHS confirm bacterial contamination.
              Treatment by most water suppliers does not meet USPHS standards.

1925         PL 68-585 authorizes 200 ft wide channel, Kansas City, Missouri, to mouth.

1927         Extension of 1.8 m deep channel to Sioux City, IA (PL 70-560).

1934         Passage of Fish and Wildlife Coordination Act (PL 73-121) requiring that fish and
              wildlife receive equal consideration to other  purposes of federal planning in federally
              funded or approved water development projects.

1936         Passage of Flood Control Act (PL 74-738) to develop "works of improvement" on more
              than 50 major river throughout the U.S.

1937         Construction completed on the first mainstem dam and impoundment on Missouri
              River, Ft. Peck Dam and Reservoir, Montana, to supply water for river navigation.

1944         Flood Control Act of 1944 (PL 78-534)  authorized Pick-Sloan Plan to construct 6 dams
              on mainstem of Missouri River.  Missouri River Bank Stabilization and Navigation
              Project authorized for flood control, bank stabilization, land reclamation, hydropower
              generation, and development and maintenance of navigation channel.

1945         Rivers and Harbors Act (PL 79-14) passed, provided a 2.7-m deep, 91.4-m wide
              navigation channel from St. Louis to Sioux City.

1946         Fish and Wildlife Coordination Act of 1946 (PL 79-732) passed, required federal
              agencies to  construct water projects with a view to preventing loss of and damage to
              wildlife resources.

1946-55       5 additional  dams and reservoirs constructed on Missouri River. See Table 4 for
              details.

1956         Federal Clean Water Act (PL 84-660) passage strengthens water quality regulations.

1958         Fish and Wildlife Coordination Act of 1958 (PL 85-624) required the cost of water
              project modifications or land acquisition earlier required under PL 79-732 to prevent
              loss or damage to wildlife be included  as part of the project costs.
                                         (continued)


                                            201

-------
Table 1. (continued)
Date
                             Event
1960-          Replacement of permeable pile dikes with impermeable rock dikes.

1960-70        Construction of primary wastewater treatment facilities for major discharges on lower
               river.

1964           Fish kill in Missouri River extending more than 161 km downstream from Kansas City,
               Missouri.

1965           Federal Water Project Restoration act (PL 89-72) required non-federal public agencies
               to administer fish, wildlife, and recreation on project lands and pay one-half of costs
               allocated to these resources.

1969           Flavor tests reveal unacceptable tastes in fishes from several locations in Missouri
               River.  PCB levels in common carp pose potential threat.

1969           Federal Water Pollution Control Authority, and later U.S. Environmental Protection
               Agency (USEPA), establishes downstream minimum  daily average flow requirements
               to maintain federally approved water quality standards.

1970-71        25% of fishes sampled from a bay in Lake Oahe, South Dakota, contained unsafe
               levels of methylmercury. Source of mercury was mining operations on a tributary
               stream.

1970-74        PCBs, aldrin, and dieldrin levels in fishes at Hermann, Missouri, pose potential health
               threat.

1971           USEPA study reveals levels of Salmonella, fecal coliform bacteria, and viruses in
               Missouri River present  a potential hazard for drinking water or recreation.

1972           Federal Water Pollution Control Act of 1972 (PL 92-500) passed requiring USEPA to
               establish national effluent and toxic discharge standards.

1972-88        Construction of secondary wastewater treatment facilities for most major dischargers
               to lower river.

1973           Endangered Species Act (PL 93-205) passed requiring U.S. Fish and Wildlife Service
               to list species threatened or endangered with extinction; authorizes programs for their
               recovery; prohibits authorization of federal projects that jeopardize listed species or
               their habitats. See Table 6 for Federally listed Missouri  River biota.

1975-80        U.S. Army Corps of Engineers (USAGE) constructs environmental notches in 1,306
               wing dikes from Sioux  City to St.  Louis to create fish habitat on downstream side of
               dike.
1976-78
PCBs, aldrin, dieldrin, and chlordane residues in fishes exceed safe limits.
                                          (continued)


                                             202

-------
Table 1. (continued)
Date
                             Event
1976
1977
1978
1980
1984-86
1986
1987-
1988
1988-90
1989
1989-
1990-
1990
Toxic Substances Control Act (PL 94-469) passed phasing out use of PCBs and
restricting use of chlordane.

Clean Water Act of 1977 (PL 95-217) established national effluent standards for water
pollutants; required cities to implement secondary sewage treatment and provided
federal grants to aid construction; set target date for discharge elimination of  129
priority pollutants; required USEPA permit for point source discharge of pollutants.

240 km of free-flowing Missouri River in Montana and 93 km below Gavins Point Dam
incorporated  into National Wild and Scenic Rivers System.

37 common industrial solvents (13 metals, 23 organic compounds, and cyanide)
detected in St. Louis water treatment plants.

Chlordane levels in fish flesh from lower Missouri River reported to exceed safe limits
for consumption.

Water Resources Development Act (PL 99-662) authorizes USACE to mitigate aquatic
and terrestrial habitat fosses from past projects.

Missouri Department of Health advisories issued warning against consumption of
specific commercial fish species from areas of Missouri River due to toxic
contamination.

Missouri River Natural Resources Committee (MRNRC) established to promote
preservation,  wise utilization, and enhancement of natural and recreational resources
of Missouri River.

Major drought in Missouri River Basin. Water shortage precipitates conflict  over water
allocations for navigation versus recreation.  USACE initiates Master Manual Review
and updates to develop and evaluate alternative water management operations for
mainstem reservoir system.

Mississippi Interstate Cooperative Resource Agreement (MICRA) formed by various
entities in the Mississippi River Basin (see text).

More  stringent discharge permit limitations for toxic metals and organics imposed on
wastewater treatment facilities of major cities along lower Missouri River, Missouri.

Upper Missouri River Basin states (Montana, N. and S. Dakota) sue USACE claiming
reservoir operation should consider upstream recreation needs in addition to lower
river navigation needs for water releases.

Missouri River Initiative (Missouri River-Conserving a River Ecosystem, MOR-CARE)
formed to facilitate cooperation among governmental, Tribal, and private parties for
optimal recovery of natural resource values and environmental health of Missouri River
ecosystem (see text).
                           (continued)
                                            203

-------
Table 1. (continued)
Date                                     Event
1990          Nonindigenous Aquatic Nuisance Prevention and Control Act of 1990 (PL 101-646)
              passed to prevent and control infestations of coastal inland waters by zebra mussel
              and other nonindigenous aquatic nuisance species.

1991          USAGE mitigation projects begin on lower Missouri River, include land purchases in
              floodplain and construction to enhance aquatic resources (see text).

1991          A 63 km section from Ft. Randall Dam to headwaters of Lewis and Clark Lake,
              including 40 km of lower Niobrara River and 13 km of Verdegie Creek, designated a
              National Recreation River

1991          Missouri Department of Conservation publishes Big River Fisheries 10 year Strategic
              Plan.

1992          Introduction of Gunderson Bill (H.R. 4169) to establish a Council on Interjurisdictional
              Rivers Fisheries and to provide funds to MICRA to conduct a comprehensive study of
              the status, management, research, and restoration needs of fisheries of Mississippi
              River drainage basin (see text).

1992          Closure of commercial fishing for all catfish species in lower Missouri River (see text).
years.  Box elder (Acer negundo), silver maple (Acer saccharinum), red mulberry (Morus
rubra), and American elm (Ulmus americana) replace cottonwood as an intermediate serial
stage.  Mature floodplain forests contain several species of oaks (Quercus spp.) and hickories
(Carya spp.), plus hackberry (Celtis occidentalis), American elm, black walnut (Juglans nigra),
green ash (Fraxinus pennsylvanica), sycamore (Plantanus occidentalis), basswood (Tilia
americana), and-almost exclusively in old growth forests-pawpaw (Asimina triloba)  (Weaver
1960, Bragg and Tatschl 1977).  Little light penetrates the dense canopy of mature Missouri
River floodplain forests resulting in an understory dominated by climbing vines (Weaver 1960),
including poison ivy (Rhus radicans), Virginia creeper (Partenocissus quinquefolia), and wild
grapes (Vitis spp.).

       Studies of the structural and functional  biology of the Missouri River abound. Sowards
and  Maxwell (1985) list more than 600 Missouri River references and Hesse et al. (1982,
1989a) provide comprehensive reviews of phytoplankton, periphyton, invertebrates,  fishes,
and  energy dynamics of the river and its mainstem reservoirs.  Many of these studies treat
how biota and processes have been influenced by river alteration, and are referenced in
Table 1.  Lower  Missouri River fish and fisheries are treated in a separate section.

ALTERATIONS TO THE MISSOURI RIVER  ECOSYSTEM

       Modifications to the integrity of the natural Missouri River-floodplain ecosystem have
been immense and ongoing for more than 150 years (Table 1).   Presently, 35% (1,316 km) of
the river's length is impounded, 32% (1,212 km) is channelized or stabilized, and the


                                          204

-------
remaining 33% (1,241 km) is free-flowing (Schmulbach et al. 1992).  Major civil works projects
included channelization, channel maintenance, and impoundment and reservoir operation.
Total cost for construction, operation, and maintenance of civil works projects through 1984
was $5.9 billion dollars (Table 2, Hesse 1987). Agricultural, industrial, and urban development
within the basin also significantly modified the Missouri River and produced extensive water
pollution (Table 1).

CHANNELIZATION

       Abundant large woody debris (snags) in the river channel, fluctuating water levels, and
extensive channel migration made early Missouri River navigation perilous.  Modifications of
the river to facilitate navigation included snag removal; channel dredging; and construction
and maintenance of dikes, revetments, and levees. Stabilizing a river channel contrasts
sharply with the concept of dynamic equilibrium discussed earlier.  Stabilized channels are
static.  They  lack the successional pattern and periodic disturbance events that maintain
physical habitat diversity.  Consequently, structure and function of the biological system  also
become stabilized.  Funk and Robinson (1974) described how channelization and associated
activities were accomplished in the lower Missouri River and we summarize its chronology in
Table 1.  Presently all of the Missouri River from Sioux City, Iowa, to its mouth at St. Louis,
Missouri, is channelized.  Even during flooding only about 10% of the original floodplain  is
inundated, as high agricultural levees confine the river to a width of 183-335 m (Schmulbach
et al. 1992).  Impacts of snag removal and channelization have been numerous and severe on
 TABLE 2. SUMMARY STATISTICS FOR CIVIL WORKS PROJECTS IN THE MISSOURI RIVER BASIN
 THROUGH 1984. CONSTRUCTION COSTS ARE ACTUAL DOLLARS SPENT, AS ARE OPERATIONS
      AND MAINTENANCE ON CORPS OF ENGINEERS PROJECTS, BUT OPERATIONS AND
    MAINTENANCE HAD TO BE ESTIMATED FOR BUREAU OF RECLAMATION PROJECTS. THE
 DOLLARS DEPICTED IN THIS TABLE ARE CONSIDERED CONSERVATIVE ESTIMATES.  (SOURCE:
                                    HESSE 1987)

                                         Millions of dollars
                 Numbers of  Construction   Operations and      Total
 Project type      projects      cost        maintenance       costs
Channel
Levee
Reservoir
Other construction
Federal recreation
Total
19
2
77
NA
19
117
561.5
92.2
3948.1
235.0
21.8
4858.6
369.0
NA
921.7
1.6
NA
1292.3
930.5
92.2
869.8
236.6
21.8
150.9
                                        205

-------
 TABLE 3.  SUMMARY OF EFFECTS OF RIVER CHANNELIZATION (C), INCLUDING SNAG REMOVAL,
 AND CONSTRUCTION OF DIKES, REVETMENTS, AND LEVEES; CONSTRUCTION AND OPERATION
 OF MAINSTREAM DAMS (D), AND BOTH TYPES OF ALTERATIONS (CD) ON THE LOWER MISSOURI
                                   RIVER ECOSYSTEM
Physical

C      Changes in channel geomorphology:
                  8% reduction in channel length
                 27% reduction in bank-to-bank channel area
                 50% reduction in original surface area of river
                 98% reduction in surface area of islands
                 89% reduction in number of islands
                 97% reduction in area of sandbars

       resulting in reduction in channel diversity through loss of side channels, backwaters,
       islands, and meandering  (Funk and Robinson 1974, Hesse et al. 1988).

C      Change in physical substrate from dominance of silt, sand, and wood to rock rip-rap.

C      Increased water depth and velocity in main channel.

D      Pre- versus post-impoundment declines in suspended sediment loads at Omaha,
       Nebraska, and St. Louis,  Missouri, of 175 to 25 and 250 to  125 million tonnes per year
       (Schmulbach et al. 1992).

D      Reduction in river sediment load resulting in channel bed degradation including
       channel deepening, increased bank erosion, and drainage  of remnant backwaters
       downstream from dams (Hesse et al. 1988, 1989a, b).

D      Silt-clay fraction of suspended sediment load reduced by 50%, but sand fraction
       increased 260%, following closure of Gavins Point Dam in 1954 (Slizeski et al. 1982).

D      Reduction in turbidity resulting  in increased light penetration (Morris et al. 1968,
       Pflieger and Grace 1987).

D      Modification of natural flow regime by evening maximum and minimum discharges and
       eliminating periodic flood pulse.

D      Reduction in annual temperature range.

CD    Loss of periodic flooding  and floodplain connectivity.
Chemical
       Higher water velocities reduce travel time for dissolved ions, nutrients, and
       contaminants.
                                    (continued)


                                         206

-------
Table 3. (continued).
 Chemical (continued)

 D      Increase in dissolved oxygen concentrations below mainstem dams (Morris et al.
        1968).

 D      Higher postimpoundment summer flows for navigation dilute impacts of point source
        discharged pollutants (Ford 1982).

 D      Reductions in nitrogen and phosphorus concentrations downstream from reservoirs
        and changes in spiraling patterns (Ward and Stanford 1983, Schmulbach et al. 1992).

 Biological

 C      Decline in habitat richness resulted in presumed decrease in diversity of periphytic
        algae (Farrell and Tesar 1982).

 C      Elimination of plankton and invertebrates produced in standing water chutes and
        sloughs with  loss of these habitats (Whitley and Campbell  1974).

 C      Loss of instream snag habitat and functions of organic matter retention and substrate
        for invertebrates and fishes (Benke et al. 1985).

 C      Greater standing crop of benthic invertebrates in main stream of unchannelized versus
        channelized river sections (Berner 1951, Morris et al.  1968, Nord and Schmulbach
        1973).

 C      Smaller standing crops of benthic invertebrates in chutes and mud banks of
        unchannelized versus  channelized sections (Morris et al. 1968).

 C      Larger standing crop of drift in unchannelized than channelized sections of river and
        little similarity between drift and benthos (Morris et al. 1968, Modde and Schmulbach
        1973).

 C      67% reduction in benthic area suitable for invertebrate colonization (Morris et al. 1968).

 C      54% decline in benthic invertebrate production from all unchannelized habitats of
        Missouri River downstream from mainstem dams between 1963 and 1980 and 74%
        decrease in production in chute/backwater habitats (Mestl and Hesse 1992).

C      Loss of river-floodplain connection for fish  migration, spawning and rearing.

C      Reduction in  microhabitats resulting in decreased abundance of fish
        species in channelized versus unchannelized section  of river in Nebraska
        (Schmulbach et al. 1975).

C      Higher standing crop of sportfishes in unchannelized  sections of river in Nebraska
       compared with channelized sections attributed to more backwater habitat and greater
       habitat diversity (Groen and Schmulbach 1978).

                                      (continued)


                                          207

-------
Table 3. (continued)
Biological (continued)

C      Loss of nesting habitat for sandbar/sand island birds leading to drastic population
       declines (e.g.,  Sterna albifrons. Charadrius melodus).

D      Elimination of riparian forests and stream channels in areas flooded by reservoirs,
       totaling more than one-third entire length of Missouri River (Hesse et al. 1988).

D      Entrainment of fluvial paniculate organic matter in reservoirs.

D      Temperature induced shifts in periphyton and phytoplankton community structure,
       particularly below dams (Farrell and Tesar 1982, Reetz 1982).

D      Increase in periphyton primary production below dams (Ward and Stanford 1983).

D      Increased relative importance of phytoplankton  biomass  and primary production
       compared with upstream  allochthonous inputs.

D      Increase in diversity and density of zooplankton community in river downstream from
       reservoirs (Repsys and Rogers 1982).

D      Changes in standing crop and diversity, and shifts in functional feeding groups of
       benthic macroinvertebrates in river downstream from reservoirs (Ward and Stanford
       1979).

D      Alteration of emergence cues, egg-hatching, diapause-breaking, and maturation of
       aquatic insects due to thermal modifications below reservoirs (Ward and Stanford
       1979, Petts 1984).

D      Blockage of riverine fish migration.

D      Inundation of floodplain fish spawning and nursery habitats.

D      Development of extensive sportfisheries in reservoirs and tailwaters (Hesse et al.
       1989a).

CD    Near elimination of natural riparian community (Hesse et al. 1988,  1989a, b).  Changes
       reported:
                   -41% deciduous vegetation
                   -12% grasslands
                   -39% wetlands

CD    25% decrease in post-dam tree growth in North Dakota floodplain  compared with
       pre-dam period related to absence of annual soil profile  saturation, lowering of water
       table in spring to reduce  downstream flooding (Reiley and Johnson 1982), and lack of
       nutrient silt deposition (Burgess et al. 1973).
                                      (continued)


                                            208

-------
Table 3. (continued)
 Biological (continued)

 CD     Increasing proportion of mature forest to other successional stages in remaining
        floodplain (Bragg and Tatschl 1977).

 CD     80% decline in organic carbon load of post-control Missouri River to Mississippi River
        compared with pre-control (Hesse et al. 1988).

 CD     Loss of major floodplain habitat types reduced populations of associated flora and
        fauna  (Clapp 1977).

 CD     Decreases in endemic large river fishes (e.g., Scaphirhvnchus albus.  Polydon
        spathula. Cycleptus elongatus. Hvbopsis gracilis) and increases in pelagic planktivores
        (e.g., Dorosoma cepedianum. Alosa chrysochloris) and sight-feeding  carnivores (e.g.,
        Morone chrysops. Lepomis macrochirus) (Pflieger and Grace 1987, Hesse et al. 1992).

 CD     Population declines of 11 native Missouri River Basin biota leading to listing as
        Federally threatened or endangered (Table 7).

 CD     As much as an 80% decline in commercial fishery in Nebraska and 97% decline in
        tailwater recreational fishery below Gavins Point  Dam (Hesse and Mestl 1992).

 CD     Decline in legal-sized catfishes in Missouri River, Missouri, attributed in part to
        increased susceptibility to exploitation due to lost habitat diversity (Funk and Robinson
        1974, Robinson 1992).

 CD     Introduction and establishment of non-native fishes and invertebrates (e.g.,
        Oncorhynchus spp., Osmerus mordax. Mvsjs relicta).  See Table 6 for list of
        introduced fishes.
Social

D     Hydroelectric power generation of over 2.2 Gw, sales totaling $1.5 billion from 1943-86
       (Sveum 1988).

D     Development of major reservoir-based recreation and associated commercial services,
       supported spending of $65 million in 1988 (General Accounting Office 1992).

CD    Commercial navigation industry transports about 2 million tonnes goods, producing
       gross revenues of $17 million in 1988 (General Accounting Office 1992).

CD    Water supply provided to 40 cities (3.2 million people), 21 power plants, and 2
       chemical manufacturers in lower Missouri River (General Accounting Office 1992).

CD    4,000% increase in area of agricultural land use  (Hesse et al. 1988).

CD    95% of protected floodplain now in agricultural, urban, and industrial uses (Hesse et
       al. 1989b).
                                            209

-------
the physical, chemical, and biological structure and function of the Missouri River and its
floodplain (Table 3). The most damaging of these alterations to aquatic communities has
been the near complete isolation of the river from its floodplain, subsequent loss of floodplain
habitat, a drastic reduction in area and diversity of river channel habitats, and an increase in
main channel flow velocity. See Brookes (1988) for a further review of the general physical
and biological impacts of river channelization.

DAM CONSTRUCTION AND OPERATION

      Widespread flooding during the war years  of 1942-44 was the impetus for passage of
the 1944 Flood Control Act to construct a six-dam mainstem Missouri River flood control
system (Keenlyne 1988). Called the Pick-Sloan Plan, it would "...provide for the most efficient
utilization of waters of the  Missouri River Basin for all purposes including irrigation, navigation,
power, domestic and sanitary purposes, wildlife, and recreation" (House Report No. 475
1944).

      The last project, Big Bend, was completed in 1963 yielding a total storage capacity for
the six reservoirs of 91.5 km3, the  largest of any system in the U.S. (Table 4).  Other large
storage reservoirs,  more than 1,300 smaller impoundments, and farm ponds also have been
built on the Missouri River mainstem and tributaries (Schmulbach et al. 1992).  Sveum (1988)
summarized the controversial operating history of the six mainstem reservoirs for their
designed multiple uses.

      Impacts of mainstem regulation on downstream lotic ecosystems are numerous and
well documented (Ward and Stanford 1979, 1983, Lillehammer and Saltveit 1984, Petts 1984,
Davies and Walker 1986, Dodge 1989, National Research Council 1992).  Reduction  in
suspended sediment loads and turbidity in the lower Missouri River has been one of the  most
obvious results of upstream impoundment (Morris et al. 1968, Whitley and Campbell 1974,
Ford 1982, Slizeski et al. 1982, Schmulbach et al.  1992).

      Average annual suspended loads decreased between 67 and 99% among various
lower river cities (Table 5)  and mean annual turbidity at the mouth of the  Missouri River above
St. Louis, Missouri, decreased fourfold from the 1930s to the 1970s (Fig.  5).  Changes in
particle sizes of suspended sediment, periphyton  growth,  and fish functional feeding groups
have all been associated with reductions in suspended load and increased water clarity
(Table 3).  Concurrently, sediment retention by mainstem reservoirs has increased the erosive
power of water discharged from dams.  The river's bed 8.3 km downstream from Gavins  Point
Dam downcut 2.3 m between 1929 and 1980 and  degradation continues to occur at least 346
km downstream to  the mouth of the Platte River (Hesse et al. 1989a).

      Alteration of the natural hydrograph has undoubtedly been the most significant impact
of dam construction to the lower Missouri River and constitutes a major disturbance (sensu
Resh et al.  1988) to the system. Bimodal March and June discharge maxima evident prior to
impoundment have been replaced by a flat hydrograph for the April through November
navigation season (Hesse  and Mestl 1992, Fig. 4). Present water management has also
reduced flushing flows or flows that exceed bank-full discharge (Hesse and Mestl 1992).
Bank-full discharge is responsible for maintaining  channel configuration and substrate
composition, and is the discharge above which the floodplain is inundated in  floodplain rivers
(Stalnaker et al. 1989). Earlier we discussed the importance of the flood  pulse to the  integrity


                                         210

-------

Q.
1
*
O
a
LU

c
^
uj C
is
Si*j
= 1
3»
CO CO
CO 00
f— III
CO >
zco
< CO
2 LU
s§
coo
9co,
Lr
g
2
i
o

3
CD
1



5 +•
lot^
l°x
IN
si.
0)


& IT
g a£
is*
Q C.

CD
CO 3 .c ^t1
^Sll

0)


fi



w
1
cc

.2 2>
o ^
o £.
-1 '^




Q 
a.
c
o
LL.

^


O)



CM

s
O)
^^


CM



CO
CD
CO
CO


CO
(\i
CM

Uj
a>
c
Garriso

«


CO
I


CO
si

§
s


CO




0>
CO
o


in
(s.
*"


ScT
2
1
O

8


CM
CO
T-



o>

cvi


i^




&
1
CO




T—
_
sf
T—
^
CD
CD
O)
CD

in


00


CO
CO

8
CD


s




'0 ®
SS O
LL.


(0
^
*~
1
^•^
«""^
T3
CO
CC
t:
o
UL

1


5


(O
in

CM
CD
0


§



^
If
so



o
CO
^
1

E
'6
Q.
CO
1
O

g
a

i
I

i
1
•5
0
i?

•5
£
3
8"

1
CD
O
CO
«
w

£ yj
S 1
^ ^f
0) Q
•^ [5
»,,
II
JC $
CC ^
* T~
211

-------
 TABLE 5. AVERAGE ANNUAL SUSPENDED SEDIMENT LOAD (MILLION METRIC TONNES) IN THE
                   LOWER MISSOURI RIVER, MODIFIED FROM FORD (1982)

                                           Sediment load
                          River                                     Percent
 Location                  (km)         <1953       >1955          change
 Yankton, South Dakota       1305         125.0         1.3             -99

 Sioux City, Iowa            1178                      10.7

 Omaha, Nebraska           1107         148.6        25.9             -83

 Nebraska City, Nebraska     904                      42.7
St. Joseph, Missouri
Kansas City, Missouri
Boonville, Missouri
Hermann, Missouri
727
579
290
161
233.3
215.9
317.5
295.9
52.3
71.7

91.4
-78
-67

-69
of large river ecosystems.  Hesse and Mestl (1992) calculated bank-full discharge (the
maximum instantaneous flow with a recurrence interval of about 1.5 years, Stalnaker et al.
1989) for the Missouri River between 1929 and 1948 as 3,115 rrvYsec. This discharge
occurred in 15 of 24 mostly predam years (1929-1952),  but in only 2 of 33 years following
closure of mainstem dams (1954-1986).  Channelization and impoundment of the Missouri
River have effectively decoupled the lower river from its  floodplain and disrupted the annual
flood pulse.  This disruption has resulted in widespread and severe disturbance to the
physical, chemical and biological character of the lower river  (Table 3).

WATER POLLUTION

       Settlement of the Missouri River floodplain was accompanied by discharge of a variety
of pollutants into the river.  As early as 1909 increases in typhoid deaths within
towns along the lower Missouri River prompted investigations into the sources of river water
pollution (Ford 1982). Significant water pollution events in the lower Missouri River include
contamination from municipal, industrial, and agricultural sources (Table 1).

       Organic pollution from untreated  human sewage, the meat packing industry, and
stockyards produces bacterial and viral contamination of drinking water supplies, sludge
deposition, and low dissolved oxygen concentrations resulting in fish kills (Ford 1982). Major
industrial pollutants in the lower Missouri River include petroleum wastes, heavy metals
(primarily mercury), and PCBs from urban industries and the processing of mined ores.
Petroleum contamination has resulted in reports of off flavors in fishes (Ford  1982), and PCB


                                         212

-------
      3000
 s>   250°
 -i  2000
 Q
 £5   1500'
 CC
 HI
      1000
       500
                      PREIMPOUNDMENT
                                     .     *   TRANSITION
                                                   .      .. POSTIMPOUNDMENT .
                                                    •.     I                   I
               1930
1940        1950        1960

               YEAR
1970
1980
Fig. 5. Mean annual turbidities (JTUs) of the Missouri River determined from daily
       measurements at the St. Louis water-treatment facility.  (Source: Pflieger and Grace
       1987.)
concentrations in fish tissue have routinely exceeded allowable federal standards (Ford 1982,
Bush and Grace 1989).  Levels of PCBs in fishes from the Missouri reach of the Missouri
River, have been declining since their manufacture and use was discontinued in 1979 (Bush
and Grace 1989). Methylation of mercury, its bioaccumulation, and biomagnification in
Missouri River tributaries, the mainstem river, and reservoirs, has created severe
environmental hazards to fish and fish-eating birds and poses a potential threat to human
health (Hesse et al. 1975, Schmulbach et al. 1992).

       Concentrations of the agricultural pesticides dieldrin and DDT occasionally violated
lower Missouri River water  quality standards during the 1970s  (Ford 1982). Violations of
federal standards for dieldrin in fish were reported during the late  1970s  (Ford 1982), but were
below advisory levels by the mid 1980s (Bush and Grace 1989). Chlordane is a pesticide
formerly used extensively in the Midwest for termite control in houses. High concentrations in
fish tissue resulted in public health advisories against consumption of selected species and
from specific river reaches  in Missouri and Nebraska (Bush and Grace 1989, Christiansen et
al. 1991).  Chlordane use was banned in 1988 and fewer fish are currently exceeding unsafe
levels (Bush and Grace 1989).  More extensive reviews of these and other contaminants in
                                        213

-------
the lower Missouri River can be found in Ford (1982) and Schmulbach et al. (1992). Water
quality legislation and enforcement requiring permits for point source pollution discharge to
rivers, use of the best available technology to control toxic pollutants, and improved
wastewater treatment for municipalities have done much to improve water quality in the lower
Missouri River (See Table 1 for a summary of important environmental legislation).

FISHES AND FISHERY RESOURCES OF THE LOWER MISSOURI RIVER

       The Mississippi River Basin, of which the Missouri River is a major tributary, supports
the richest freshwater fish fauna in North America, about 260 species (Robison 1986).
Freshwater dispersants make up about 88% of these species (Moyle and Cech 1988),
predominantly Cyprinids  (30%), Percids (26%), Centrarchids (7%), Catostomids (9%), and
Ictalurids (5%); while diadromous (7%) and freshwater representatives of marine families (5%)
are poorly represented. The Mississippi  River Basin was also uniquely important in North
America as a center of fish evolution, as  a refuge during times of glaciation from which
species have been able to reoccupy water vacated during glacial advances and as a refuge
of ancient fish faunas (Moyle and Cech 1988). The archaic families Acipenseridae,
Polyodontidae, Lepisosteidae,  and Hiodontidae are all extant in the Missouri River Basin.

       Hesse et al. (1989a) reviewed the fishes and fisheries of the entire Missouri River
Basin.  We will, therefore, concentrate on selected aspects of the lower basin.  Ninety-one fish
species have been reported from the mainstem Missouri River, Missouri (Table 6, Grace and
Pflieger 1989).  Surveys made  at approximately 20-year intervals from 1940 to 1983 in  the
Missouri reach were analyzed  by Pflieger and Grace (1987) and show an increase  in the
number of species collected and substantial changes in their relative abundances.  Species
reported to become established or more abundant were mostly pelagic planktivores and
sight-feeding carnivores:  skipjack herring, gizzard shad, white bass, bluegill, white crappie,
emerald shiner, and red shiner (see  Table 6 for scientific names).  These shifts appear related
to decreased turbidity (Fig. 5)  and changes in the flow regime following impoundment.

       Fishes that declined over the same period included common  carp, river carpsucker,
bigmouth buffalo, and two endemic  large river species-pallid sturgeon and flathead chub.
The pallid sturgeon was never reported as abundant throughout its range in the
Missouri-Mississippi River Basin (Kallemeyn 1983). However,  dangerously low populations
and hybridization with shovelnose sturgeon (Carlson et al. 1985) posed such threats to the
species survival that it was listed in 1990 as endangered under the Endangered Species Act
(Table  7,  16 U.S.C. 1531). Habitat alteration and destruction due to dam construction  and
channelization, as summarized in Table 3, are cited as major factors responsible for this
species' decline (Deacon et al. 1979, Kallemeyn  1983, Pallid Sturgeon Recovery Team 1992).

       Reported declines in numbers of other mainstem Missouri River fishes in the Nebraska
reach include several large species:  paddlefish, sauger, flathead catfish, and blue sucker; in
addition to several chub species: flathead, speckled, sturgeon, sicklefin, and silver  (Hesse
and Mestl 1992). Similar reductions in populations of small fishes were reported  by Pflieger
and Grace (1987) in the uppermost sections of the Missouri reach, but populations in
lowermost sections of the Missouri reach appear stable or increasing.  Populations of  two
silvery  minnows (western silvery minnow and plains minnow), which typically occur in
backwater habitats, have also  declined throughout the lower Missouri River because of
habitat loss (Pflieger and Grace 1987, Hesse and Mestl 1992).

                                        214

-------
   TABLE 6. FISH FAMILIES AND SPECIES OF THE MISSOURI RIVER, MISSOURI, INCLUDING
 PRESENT FEDERAL (CANDIDATE FOR LISTING: C2, THREATENED: FT, ENDANGERED: FE) AND
 MISSOURI (WATCH LIST: WL, RARE: R, ENDANGERED: SE) STATUS AND IF INTRODUCED (I) TO
      THE BASIN. (SOURCES: GRACE AND PFUEGER 1989, MISSOURI DEPARTMENT OF
                                CONSERVATION 1991)
Family and species                                                  Status
Petromyzontidae
      Chestnut lamprey, Ichthvomvzon castaneus

Acipenseridae
      Lake sturgeon, Acipenser fulvescens                              C2, SE
      Shovelnose sturgeon, Scaphirhvnchus platorynchus
      Pallid sturgeon, Scaphirhvnchus albus                            FE, SE

Polyodontidae
      Paddlefish, Polyodon spathula                                   WL

Lepisosteidae
      Shortnose gar, Lepisosteus platostomus
      Longnose gar, Lepisosteus osseus

Anguillidae
      American eel, Anquilla rostrata

Clupeidae
      Skipjack herring, Alosa chrvsochloris
      Alabama shad, Alosa alabamae                                  R
      Gizzard shad, Dorosoma cepedianum

Hiodontidae
      Goldeye, Hiodon alosoides
      Mooneye, Hiodon tergisus                                      R

Osmeridae
      Rainbow smelt, Osmerus mordax                                 I

Esocidae
      Northern pike, Esox lucius

Cyprinidae
      Common carp, Cyprinus carpio                                   I
      Grass carp, Ctenopharyngodon idella                             I
      Bighead carp, Hypophthalmichthvs nobilis                          I
      Silver carp, Hypophthalmichthys molitrix                            I
      Golden shiner, Notemigonus crysoleucas
      Creek chub, Semotilus atromaculatus
      Silver chub, Hybopsis storeriana

                                     (continued)


                                       215

-------
TABLE 6. (continued)
 Family and species                                                     Status
 Cyprinidae (continued)
       Gravel chub, Hvbopsis x-punctata
       Speckled chub, Hvbopsis aestivalis                                 SE
                                                                        Flathead chub,
                                                                        Hvbopsis gracilis
       Sturgeon chub, Hvbopsis gelida                                    C2, R
       Sicklefin chub, Hvbopsis meeki                                     C2, R
       Suckermouth minnow, Phenacobius mirabilis
       Emerald shiner, Notropis atherinoides
       Rosyface shiner, Notropis rubellus
       Western redfin shiner, Notropis umbratilis umbratilis
       Silverband shiner, Notropis shumardi
       Common shiner, Notropis cornutus
       Striped shiner, Notropis chrysocephalus
       River shiner, Notropis blennius
       Bigeye shiner, Notropis boops
       Bigmouth shiner, Notropis dorsalis
       Spotfin shiner, Notropis spilopterus
       Red shiner, Notropis lutrensis
       Sand shiner, Notropis stramineus
       Channel mimic shiner, Notropis wickliffi
       Ghost shiner, Notropis buchanani                                   WL
       Brassy minnow, Hvbognathus hankinsoni                            R
       Western silvery minnow, Hvbognathus argyritis
       Plains minnow, Hvbognathus placitus
       Bluntnose minnow, Pimephales notatus
       Fathead minnow, Pimephales promelas
       Central stoneroller, Campostoma anomalum
       Largescale stoneroller, Campostoma oliqolepis

 Catostomidae
       Blue sucker, Cycleptus elongatus                                   C2, WL
       Bigmouth buffalo, Ictiobus cyprinellus
       Black buffalo, Ictiobus niger
       Smallmouth buffalo, Ictiobus bubalus
       River carpsucker, Carpiodes carpio
       Highfin carpsucker, Carpiodes velifer                                R
       Quillback, Carpiodes cyprinus
       White sucker, Catostomus commersoni
       Northern hog sucker, Hypentelium nigricans
       Golden redhorse, Moxostoma erythrurum
       Shorthead redhorse, Moxostoma macrolepidotum

 Ictaluridae
       Black bullhead, Ictalurus melas
       Yellow bullhead, Ictalurus natalis

                                         (continued)


                                            216

-------
TABLE 6. (continued)
Family and species                                                       Status
 Ictaluridae (continued)
       Channel catfish, Ictalurus punctatus
       Blue catfish, Ictalurus furcatus
       Tadpole madtom, Noturus gyrinus
       Freckled madtom, Noturus noctumus
       Stonecat, Noturus flavus
       Flathead catfish, Pylodictis olivaris

 Gadkjae
       Burbot, Lota lota

 Cyprinodontidae
       Blackstripe topminnow, Fundulus notatus

 PoecilikJae
       Mosquitofish, Gambusia affinis

 Atherinidae
       Brook sitverside, Labidesthes sicculus

 Percichthyidae
       White bass, Morone chrvsops                                      I
       Striped bass, Morone saxatilis                                      I
       Hybrid striper, Morone chrvsops x Morone saxatilis                   I

 Centrarchidae
       Spotted bass, Micropterus punctulatus
       Smallmouth bass, Micropterus dolomieu
       Largemouth bass, Micropterus salmoides
       Warmouth, Lepomis gulosus
       Green sunfish, Lepomis cyanellus
       Orangespotted sunfish, Lepomis humilis
       Longear sunfish, Lepomis megalotis
       Bluegill, Lepomis macrochirus
       Rock bass, Ambloplites rupestris
       White crappie, Pomoxis annularis
       Black crappie, Pomoxis nigromaculatus

 Percidae
       Walleye, Stizostedion vrtreum vitreum
       Sauger, Stizostedion canadense
       Slenderhead darter, Percina phoxocephala
       Ozark logperch, Percina caprodes fulvitaenia
       Johnny darter, Etheostoma nigrum

 Sciaenidae
       Freshwater drum, Aplodinotus grunniens
                                            217

-------
       Enrichment of lower Missouri River fish species diversity during the past 40 years
appears due largely to accidental (e.g., Asian carps) and intentional (e.g, rainbow smelt,
Morone spp.) introductions (Table 6), and an increased frequency of species in the
mainstem that are stragglers from tributaries (Pflieger and Grace 1987).  Regulation of the
river appears to have reduced environmental constraints on these species.

COMMERCIAL AND RECREATIONAL FISHERY IN THE MISSOURI REACH

       Data have been compiled since 1945 by the Missouri Department of Conservation
(Robinson 1992) for the Missouri River, Missouri, on the number of licensed commercial
fishers (determined from permit sales), weight of their reported catches, and amount and type
of gear used. Numbers of commercial fishers gradually decreased from 1948 to 1963,
remained fairly stable through 1969 and then increased to a peak of 1,039 in 1982 (Fig. 6).
Causes for these fluctuations are unknown.  However, the subsequent decline in permit sales
        TABLE 7.  FEDERALLY LISTED CANDIDATE (C1, C2, C3), THREATENED (T) AND
    ENDANGERED (E) SPECIES ENDEMIC TO THE MISSOURI RIVER FLOODPLAIN MISSOURI.
                       (SOURCE: WHITMORE AND KEENLYNE 1990)

  Species                                                    Status

 Plants
       Western prairie fringed orchid, Platanthera praeclara           T

 Insects
       American burying beetle (Nicrophorous americanus)           E
       Regal fritillary butterfly (Speyeria idalia)                      C2
       Six-banded longhorn beetle (Drvobius sexnotatus)             C2

 Fish
       Listings included in Table 6

 Reptiles
       Alligator snapping turtle (Macroclemys temminckii)             C2

 Birds
       Interior least tern (Sterna antillarum)                         E
       Piping plover (Charadrius melodus)                         T
       Whooping crane (Grus americana)                          E
       Bald eagle, Haliaeetus leucocephalus                       E
       Peregrine falcon, Falco peregrinus                          E
       Swainson's hawk (Buteo swainsoni)                         C3
       Eskimo curlew (Numenius borealis)                         E
       Migrant loggerhead shrike (Lanius migrans)                   C2
       Swallow-tailed kite  (Elanoides forficatus)                     C3

 Mammals
       Gray bat (Myotis grisescens)                               E
       Indiana bat (Myotis sodalis)                               E
                                         218

-------
                       45  SO   55  60  65   70  75  80  85  90
        1945
1950
1955
1960
1965     1970
   Years
1975
1980
1985
1990
Fig. 6.  Number of commercial fishers from the Missouri River, Missouri, and their reported
       harvest of all fish species and catfish species, 1945-1990.
from 1982 to 1988 may be due to increased commercial fishing fees implemented in 1984,
and health advisories issued from 1987 to 1990 warning against consumption of Missouri
River fishes (Robinson 1992).  Total harvest first peaked in 1945 at 228 mt, then declined
gradually to 35 mt in 1966, paralleling the decrease in numbers of fishers (Fig. 6).  Methods of
estimating annual harvest changed in 1967, yielding a more accurate but higher reported
harvest. Methods of estimating harvest have been constant since 1975 and harvest has
gerierally increased  since then, with 1989 and 1990 being the highest years recorded (Fig. 6).
These record harvests occurred despite the dramatic decline in numbers of fishers observed
in the mid 1980s. Reasons for these high catches vary, but include low river water resulting
from a basin-wide drought, increased vulnerability to gear, more accurate reporting of catch,
and lack of concern over successive health warnings (Robinson 1992). The most important
species in the commercial catch are common carp, buffalo fishes, and catfishes. Additional
species contributing to the commercial harvest include:  freshwater drum, carpsuckers,
paddlefish, and shovelnose sturgeon.

      Catfish harvest, particularly channel catfish, has increased dramatically during the past
10 years (Fig. 6).  Concomitantly, there has been a continuous decline in the proportion of
                                        219

-------
    16-r
                            — Channel Catfish     -  Flatheod Catfish
     1980    1981    1982    1983    1984    1985   1986   1987   1988    1989    1990    1991
Fig. 7. Percentage of legal length (>381  mm total length) channel and flathead catfishes
       captured by 2.5 cm mesh hoopnets (channel catfish) and electrofishing (flathead
       catfish) from Missouri River, Missouri, August-November 1990.
legal (>381 mm total length) channel and flathead catfishes in research harvests from
Missouri (Fig. 7) and other states bordering the lower Missouri River.  Record high
commercial catfish catches and a shift in population size structure to sublegal lengths implies
overharvest. Consequently, all states bordering the lower Missouri River have recently
prohibited commercial catfish harvest.

       An additional impetus for closure of the lower Missouri River commercial catfish fishery
was that most of the commercial  harvest was captured  by very few fishers while Missouri
River recreational fishing is ranked as the  number one public activity on the river (Weithman
and Fleener 1988).  Analysis of the number of commercial fishers and their reported harvest
shows that 84% reported total annual catfish catches of less than 230 kg, while only 3%
reported catches over 2000 kg.  Nearly 50 and 70% percent of the total 1990 Missouri reach
commercial catfish catch of 132,120  kg was reported by 10 and 25 fishers, respectively (Fig.
8). Weithman and Fleener (1988) estimated recreational fishing  on the Missouri reach to be
86,000 days per year (96 visits/km or 3.1 visits/ha) from 1983 to 1985.  Seventy percent of this
effort was for catfishes, contributing 57% of the total catch (212,000 fish) and 69% of the total
                                          220

-------
    14000
    12000
    10000
  in 8000
    6000
    4000
    2000
          1   2  3  4  5  6   7   8   9  10  11  12  13  14 15  16  17  18  19  20  21  22 29 24 25
                                         Fisherman Rank
Fig. 8. Rank of the top 25 fishers and their reported total catch of all catfish species from
       Missouri River, Missouri, 1990.
fish harvested by the recreational fishery.  Net annual economic benefits of recreation were
estimated at $1.9 million and recreational fishing at $660,000 on the Missouri reach. These
totals compare with an estimated net wholesale value of the commercial catch for all species
from the Missouri River of $128,000 in 1987 (Robinson 1989). Future recreational worth of the
Missouri River is perceived by resource professionals as the highest of Missouri's six
watersheds (Bachant et al. 1982).  Closure of the lower Missouri River commercial catfish
fishery should allocate a greater proportion of this resource to a larger number of citizens and
yield greater economic benefit to the area.

RESTORATION OF THE LOWER MISSOURI RIVER

       Restoration is defined by the National  Research Council (1992) as the  return of an
ecosystem to a close approximation of its condition prior to disturbance.  This definition
necessitates that both ecosystem structure and function be recreated.  Merely recreating the
form without the functions, or the functions in an artificial configuration with little resemblance
to the natural river, does not constitute restoration (National Research Council 1992).
Restoration must be conducted as a holistic process at the landscape scale of the river basin.
                                          221

-------
It cannot be achieved through isolated manipulation of individual structural or functional
elements or river segments.

       The essence of a fluvial system is the dynamic equilibrium of the physical system,
which in turn establishes a dynamic equilibrium in the biological system (National Research
Council 1992).  The goal of fluvial restoration, therefore, is reestablishment of this dynamic
equilibrium.  The National Research Council's (1992) report on Restoration of
Aquatic Ecosystems lists four general objectives for fluvial restoration that can be applied to
the Missouri, Volga, or any large river.  These objectives are:

       1.  Restore the natural water and sediment regime.  Time scales of daily-to-seasonal
variations in water and sediment loads and the annual-to-decadal patterns of floods and
droughts constitute this regime.

       2.  Restore the natural channel geometry, if restoration of the water and sediment
regime does not.

       3.  Restore the natural riparian plant community, if the natural plant community does
not restore itself upon completion of objectives 1 and 2.

       4.  Restore native aquatic plants and animals if they do not recolonize on their own.

       Restoring the  natural water and sediment regimes and the dynamic equilibrium of the
Missouri and Volga rivers is a tremendous challenge, given their size,  present state of
alteration, competing uses for water, existing water and channel control structures, and
floodplain development.  However, as acknowledged by the National Research Council
(1992), fluvial restorations are exercises in approximation, and restoration  of the Missouri and
Volga rivers will only be successful through compromise among competing interests.

       If restoration of the Missouri River is to  succeed, the four broad objectives listed above
must be realized.  We believe this goal can be accomplished by the groups  described below
through coordinating management of all significant ecological elements at a comprehensive
river basin scale, termed integrated aquatic ecosystem restoration  (National  Research Council
1992).  Restoration should include both nonstructural methods (that do not involve physical
alteration or building of structures) as well as traditional structural techniques (National
Research  Council 1992), and follow specific approaches recommended by Hesse et al.
(1989a, 1989b), Hesse and  Mestl (1992),  and Hesse et al. (1992).  We summarize these
approaches as six Missouri  River restoration objectives and recommend a range of strategies
to attain them.

Objective  1. Reestablish a semblance of the precontrol natural hydrograph

Strategies-

       1.1.  Hesse and Mestl (1992) suggest timing reservoir releases to emulate the
precontrol hydrologic cycle  as a daily percentage of total annual discharge (Fig.  9).  They
contend this approach would recreate the timing of the historic March and June  rises while
minimizing flooding and incorporate flexibility for drought or wet years.
                                          222

-------
       1.2. Controlled flooding is necessary to reestablish the river-floodplain linkage and
maintain habitat diversity. True river restoration must take discharge dynamism into account
by allowing enough spatial and temporal scope for flooding to occur (National  Research
Council 1992).  The capability for controlled flooding through reservoir releases currently
exists.

       1.3. Deauthorize irrigation projects of the Pick-Sloan Plan. Water supply in the
Missouri River Basin is insufficient to meet the extensive development envisioned by this plan
and poses the greatest threat to current water uses (Lord et al. 1975).

Objective 2.  Reestablish a semblance of the precontrol sediment regime

Strategies-

       2.1. Reduce sediment entrapment in reservoirs. Hesse et al. (1989a) and Singh and
Durgunoglu (1991) suggest methods to bypass sediment through reservoirs, which if
implemented or modified for Missouri River reservoirs would benefit  the reservoirs by
improving water quality, increasing hydropower potential, and increasing reservoir storage life.
Sediment bypass would benefit the river by  reducing downstream channel bed degradation,
providing material for restoration of natural channel geometry and habitats, contributing
sediment and organic matter to downstream reaches, and favoring native turbid-water fishes
over sight-feeding carnivores.

Objective 3.  Restore some of the  precontrol channel structural diversity and the
river-floodplain linkage

Strategies-

       3.1. Accomplish Objectives 1 and  2. This will do much to meet Objective 3.

       3.2. Implement a variety of nonstructural floodplain management methods that
promote floodplain restoration, including:

             3.2.1. Congressional establishment of U.S. Army Corps of Engineers'
       reservoir operating priorities based on economic, environmental, social, and other
       benefits to be derived from all authorized project purposes (General Accounting
       Office 1992);

             3.2.2. Adoption of regulatory  floodway zones  (Brookes 1988),  purchase of
       easements to prevent construction in the floodplain, increase land acquisition by
       willing purchase, buy out of drainage or levee districts,  easements, and  fee-title
       acquisition;

             3.2.3. Support changes in flood damage insurance policies to discourage
       continued urban and industrial floodplain development;

             3.2.4. Reduce public financial support for private sector economic gain
       by allocating maintenance costs of channel regulatory structures to direct
       beneficiaries (i.e., navigation companies and drainage and levee districts).


                                         223

-------
      Jon
                                        Jun    Jut
                                          Month
Nov
Dec
Fig. 9. Percentage of annual discharge for the Missouri River at Boonville, Missouri (Rkm
       317), by month for the preregulation years 1929-1948. This is an example of
       the natural hydrograph approach Hesse and Mestl (1992) advocate to restore the
       natural water regime of the lower Missouri River.
       3.3. Structural techniques for channel restoration should stress low-cost, low
maintenance soft engineering (recreation of the natural river channel geometry using fluvial
geomorphic principles and locally available materials) over traditional hard hydraulic
engineering approaches which use concrete, rip-rap, and other imported materials and have
high maintenance costs.  See Brookes (1988) and National Research Council (1992) for
additional structural techniques.

       3.4. Expand the Missouri River Mitigation  Project to include the entire length of
existing river channel and mainstem and tributary impoundments,  and have it incorporate an
integrated aquatic ecosystem approach to mitigation.

Objective 4. Reestablish and enhance native Missouri River fishes and their migrations

Strategies-
                                         224

-------
       4.1. Construct fish bypasses or elevators at selected dam barriers.

       4.2. Modify reservoir releases of water to provide precontrol thermal cues including
synchronization of river stage-temperature relationships.

       4.3. Enhance the present species-centered recovery efforts by incorporating fish
guild/community and habitat based approaches.

       4.4. Implement a more equitable balance between management for river recreational
fisheries and native species restoration.

       4.5. Implement more restrictive commercial and sport fishery harvest regulations
where populations of native fishes show declining trends (e.g., sauger, blue sucker, and other
species in Nebraska, Hesse et al. 1992).

Objective 5. Reduce or  eliminate major point and nonpoint sources of pollution

Strategies-

       5.1. Improve enforcement of existing point-source water quality regulations.  Establish
long-term goal of upgrading water quality of the lower Missouri River to meet criteria for
whole-body contact.

       5.2. Continue to expand governmental support for agricultural land use practices that
reduce soil erosion and  crop overproduction and conserve or enhance wetlands (e.g.,
Conservation Reserve and Wetlands Reserve programs).

       5.3. Incorporate  a multi-level water quality restoration strategy, including elements of
isolation, removal, transfer, and dilution through space and time (Herricks and  Osborne 1985).

Objective 6.  Reestablish native terrestrial and wetland plant communities along the river
channel and floodplain. including native prairie, wet meadow, bottomland hardwood forests.
sandbar, and sand island successional plant communities and vegetated islands (Hesse et al.
1989a)
Strategies-

       6.1. Remnants of native plant communities exist and should become reestablished
upon achievement of Objectives 1 -3.

RESTORATION INITIATIVES AFFECTING THE MISSOURI RIVER

       After decades of degradation, there is now a flurry of restoration activity for Missouri
River Basin natural resources.  Current initiatives, strategies, and action plans, if implemented,
have the potential to greatly enhance lower Missouri River aquatic resources. We summarize
several of these to illustrate how U.S. governmental and administrative processes operate in
the area of environmental management. Our attention as ecologists is often focused on the
biological details of our disciplines. However, we must be equally cognizant of the policy
                                         225

-------
aspects of river management, because it is here that the decisions are made and the money
allocated which enable restoration to occur.

Mississippi Interstate Cooperative Resource Agreement (MICRA)

       Large river drainage basins seldom exist within a single political boundary; rather, they
typically are  interjurisdictional. Interjurisdictional rivers are defined as crossing or common to
two or more state, provincial, country, etc., boundaries and coming under the shared
jurisdiction of two or more governmental entities.  Restoration of interjurisdictional rivers is
hampered by the multiplicity of authorities responsible for their management.  The MICRA is
an example of  an attempt at a unified approach to restoration of the Mississippi  River and its
over 90 interjurisdictional tributaries. Twenty-eight State conservation departments having
fisheries jurisdiction  in the  Mississippi  River drainage system, plus the American  Fisheries
Society and  U.S. Fish and  Wildlife Service (USFWS), have invited other federal, nonfederal,
tribal, and private entities to band together, 'To assess the Mississippi River drainage fishery
resources and  habitat requirements to protect, maintain, and enhance interstate  fisheries in
the basin" (Rasmussen 1991). The Mission of MICRA is to "Improve the conservation,
development, management, and utilization of interjurisdictional fishery resources in the
Mississippi River Basin through improved coordination and communication among the
responsible management  entities" (Rasmussen 1991).  MICRA is managed by an interagency
Steering Committee  composed of personnel from member states and entities. They
completed a Comprehensive Strategic Plan in August 1991 (Rasmussen 1991) detailing step-
by-step goals and objectives to complete MICRA's mission.  Goals set by MICRA are
presented here as an example of a river basin-wide restoration effort.

       A. Develop a formal framework and secure funding for basin-wide networking and
coordination mechanisms that complement existing and emerging administrative entities.

       B. Develop public information and education programs to disseminate information that
supports fishery resource management in the Mississippi River Basin.

       C.  Develop an information  management program based on standardized methods for
collecting and  reporting fishery resource data basin-wide.

       D. Determine and document the socio-economic value of fishery resources and
related recreation.

       E. Improve communication and coordination among entities responsible for fisheries
resources management in the Mississippi River Basin.

       F. Periodically identify and prioritize issues of concern in the Mississippi  River Basin
for coordinated research that supports cooperative resource management.

       G.  Identify and coordinate fishery management programs to address species and
habitat concerns from an ecosystem perspective.

       H. Develop compatible regulations and policies for fishery management  to achieve
interstate consensus on allocation  of fishery resources.
                                          226

-------
       I. Develop protocols, policies, and regulations for disease control, introduction of
exotics, maintenance of genetic integrity, and maintenance and enhancement of indigenous
species.

       J.  Preserve, protect, and restore fishery habitats basin-wide.

       MICRA is seeking federal funding to accomplish these goals through the Cooperative
Interjurisdictional  Rivers Fisheries Resources Act of 1992.

Cooperative Interjurisdictional Rivers Fisheries Resources Act of 1992 (H.R. 4169)

       A nationwide consensus is emerging that construction and maintenance of waterway
developments are responsible for much of the decline of large river natural resources in the
U.S, and the public has become increasingly aware  of the need for restoration of the river-
riparian ecosystem (National Research  Council 1992). Moreover, they recognize that without
demonstrable changes in current management strategies there will be further loss of
large-river fisheries and  reduced opportunities for recreational, commercial, subsistence, and
aesthetic uses of our river-floodplain ecosystems.  Several programs have been proposed or
are currently underway, including those described herein, to resolve conflicts among
management strategies.  The Cooperative Interjurisdictional Rivers Fisheries Resources Act of
1992 was introduced to the U.S. Congress to improve coordination, cooperation, research,
and information sharing at the national  level on the variety of present  programs to conserve
fisheries resources of major U.S. interjurisdictional rivers.

       If approved, this  bill will fund and establish for 3  years a Council on Interjurisdictional
Rivers Fisheries (Council) and authorize a pilot test of MICRA. Funds requested include $1
million per year to support Council activities and $2  million per year for MICRA. Membership
on the Council would include high-level representatives of federal and state agencies with
interests in fisheries resources on interjurisdictional rivers. The Council is to develop
strategies on the  management of interjurisdictional rivers fisheries.  These strategies would
include a) listing the 10  highest-priority  interjurisdictional rivers in need of cooperative fisheries
management and b) development of comprehensive fishery strategic plans for the five
highest-priority interjurisdictional rivers identified in a).

       The pilot test of MICRA would include nine tasks. Briefly these tasks are

       1.  Identification and description of each of the river ecosystems  in the Mississippi
River Basin and their associated fishery resources and habitat.

       2.  Identification and description of impacts of, and mitigation for, water and waterway
development projects on fishery resources.

       3.  Analysis of existing data on regional depletion of important fish stocks and the
potential for their  restoration.

       4.  Identification of major information gaps and technological needs to improve the
cooperative management of interjurisdictional fisheries resources.
                                          227

-------
       5. A comprehensive study of the status, management, research, and restoration
needs of the interjurisdictional fisheries of the Mississippi River Basin.

       6. Development of recommendations for MICRA participants to undertake cooperative
management and research projects.

       7. Development of plans and projects for restoration and enhancement of depleted
fish stocks and their habitats.

       8. Evaluation of MICRA and the merits of extending such a program to other river
basins in the U.S.

       9. Estimates of funds required  to implement 6, 7, and 8.

MISSOURI RIVER INITIATIVE

       A specific example of a Mississippi River subbasin effort to address management
strategies is the USFWS's proposed Missouri River Partnership or MOR-CARE (Missouri River-
-Conserving a River Ecosystem). The  goal of MOR-CARE would be similar to MICRA's, but is
restricted to the Missouri River Basin.  MOR-CARE has four draft objectives (Brabander 1992):

       1. To facilitate establishment and coordination of an operational Missouri River
environmental research, management,  restoration, and enhancement program involving
federal, state, tribal, and local governments and public interest groups.

       2. To coordinate the preparation, facilitation, and implementation of a comprehensive
Action  Plan for the management, restoration, and enhancement of fish, wildlife, and related
environmental and recreational resources within the Missouri River ecosystem in concert with
existing and future navigation, flood control, and water supply needs.

       3. To develop and implement plans for providing fish and wildlife resource-based
recreational opportunities for the people of the Missouri River ecosystem.

       4. To establish a functional outreach program  to involve and exchange information
with the public concerning problems, opportunities, and resource management and
restoration needs in the Missouri River ecosystem.

       MOR-CARE would operate through a Steering Committee organized similarly to
MICRA. This Steering Committee would identify resource management problems and
information  needs and establish Working Groups to develop information and alternatives for
problem solution. Products of the Working Groups would be incorporated by the Steering
Committee into a Missouri River Plan of Action. This Plan would include a Cooperative
Agreement for signature by MOR-CARE partners indicating all would agree to utilize their
resources to accomplish the goals of the Plan.
                                        228

-------
MISSOURI RIVER FISH AND WILDLIFE MITIGATION

       The National Research Council (1992) defines mitigation as, "...actions taken to avoid,
reduce, or compensate for the effects of environmental damage."  Possible mitigation actions
include restoration, enhancement, creation, or replacement of damaged ecosystems.

       Channelization of the Missouri River directly eliminated 40,591 ha of aquatic habitat
and 151,479 ha of wetlands and terrestrial habitat from the river and its floodplain from Sioux
City, Iowa, to St. Louis, Missouri (Hesse et al. 1989b). The initial attempt to mitigate for these
habitat losses was Section 601  (a) of the Water Resources Development Act of 1986 (PL 99-
662), which authorized the Missouri River Fish and Wildlife Mitigation Project within the states
affected by river channelization  (Missouri, Kansas,  Iowa and Nebraska).

       Estimated cost of the Missouri River Mitigation Project between 1990 and 1999 is
$67.7 million.  Funds are to be allocated among four main project elements (USAGE 1992): 1)
48% of requested funds will be for habitat development of 12,100 ha of newly acquired
nonpublic lands and 7,365 ha of existing public lands; 2) 44% will be for acquisition of the
nonpublic lands to provide aquatic and terrestrial habitat for fish and wildlife; 3) 5% will be for
planning, engineering, and design of the project, including 0.4% for baseline  evaluation and
monitoring; and 4) 3% will be for construction management.

       The Missouri River  Mitigation Project will acquire 6.3% of the  habitat lost in the original
floodplain (Hesse et al. 1989b),  a small beginning toward restoration of the Missouri River
ecosystem. The authorization plan recognizes that restoration of the Missouri River is a long-
term goal  rather than short-term objective (Hesse et al. 1989b).  This acknowledgment is
important  because this first mitigation step displays a piecemeal strategy, an example of,
"...the isolated manipulation of individual elements approach," (National Research Council
1992) and therefore insufficient by itself to achieve restoration.
                                    CONCLUSIONS
       Fewer than 200 years have elapsed since Lewis and Clark encountered the abundant
natural resources of the Missouri River Basin.  Europeans have a much longer history of river
modification than we in the U.S.  Consequently they also are more familiar with river
restoration and we should look to them for guidance (Petts 1984,  Brookes 1988, Dodge
1989). The days of the truly Big Muddy are history. Missouri River restoration should be
directed toward reestablishment of a basin-wide dynamic equilibrium and maintenance of
natural functions and characteristics, albeit in a more limited scope.

       Missouri River mitigation does not yet embody this holistic view toward restoration of
the structural and functional attributes of the unaltered river on the landscape scale of the
entire basin.  This is not to say that current small-scale restoration efforts are ineffective.
However, success in recreating a self-sustaining Missouri River ecosystem is more probable if
individual mitigation and restoration projects are planned within the context of the basin
landscape. Decisions about restoration and management of aquatic resources should not be
made on a small-scale, short-term, site-by-site basis, but should be made to promote the
long-term sustainability of all aquatic resources (National  Research Council 1992).  The task


                                         229

-------
of such groups as the Council on Interjurisdictional Rivers Fisheries, MICRA and MOR-CARE's
Steering Committee will be to provide this perspective.
                                ACKNOWLEDGMENTS
       Views expressed in this paper reflect the opinions of the authors and do not
necessarily represent policies of their agencies.  We thank J. Brabander, J. Bush, G. Farabee,
K. Keenlyne, M. LaValley, W.  Pflieger, J. Rasmussen, R. Sparks, and N. Stucky for reviewing
this manuscript.  Dennis Figg, Missouri Department of Conservation (MDC), provided
information on federal listings of Missouri River biota.  Brenda Gary, MDC prepared some of
the figures; and Sandy Clark, Missouri Cooperative Fish and Wildlife Research Unit, edited the
manuscript.  This is a contribution from the Missouri Cooperative Fish  and Wildlife Research
Unit (U.S. Fish and Wildlife Service, Missouri Department of Conservation, University of
Missouri, and Wildlife Management Institute cooperating).
                                        230

-------
                                    REFERENCES
Bachant, J. P., D. J. Witter and J. D. Martindale. 1982.  Ranking the recreational values
       and associated problems of Missouri's major watersheds.  The Regional Watershed
       Assessment Plan.  Missouri Dept. Conservation, Jefferson  City.

Bayley, P. B.  1991. The flood pulse advantage and the restoration of river-floodplain
       systems.  Regulated Rivers 6:75-86.

Benke, A. C., R. L  Henry, D. M.  Gillespie and R. J. Hunter. 1985.   Importance of snag
       habitat for animal production in southeastern streams.  Fisheries 10(5):8-13.

Benson, N. G. (ed). 1988. The  Missouri River-the resources, their uses and values.
       North  Central Div., Am. Fish. Soc., Spec. Pub. No.  8.

Berner, L M.  1951. Limnology  of the lower Missouri River.  Ecology 32:1-12.

Biette, R. M.,  G. A.  Goodchild and S. J. Nepszy. 1989.  Science networks for large river
       management. Pages 600-606 in D. P. Dodge (ed), Proc. international large river
       symposium. Can. Spec.  Publ. Fish. Aquat. Sci. 106, Dept.  Fisheries and Oceans,
       Ottawa.

Brabander, J.  1992.  MOR-CARE. What is this animal? (Draft). U.S. Fish and Wildlife
       Service, Columbia Field Office, Columbia, Missouri.

Bragg, T. B. and A. K. Tatschl.  1977.  Changes in flood-plain vegetation and land use
       along  the Missouri River from 1826 to 1972.   Environ. Manage. 1:343-348.

Brinson, M. M., A. E. Lugo and S. Brown.  1981. Primary productivity, decomposition and
       consumer activity in freshwater wetlands. Ann. Rev. Ecol. Syst.  12:123-161.

Brookes, A.  1988.  Channelized rivers-perspectives for environmental management.
       John Wiley  & Sons, Chichester.

Burgess, R. L, W. C. Johnson and W. R. Keammerer.  1973.  Vegetation of the Missouri
       River floodplain in  North  Dakota. Technical Completion  Report, Office of Wat. Resour.
       Res. Proj. No. A-022-NDAK.

Bush, J. and T. B. Grace.  1989.  Fish contaminant study of the Missouri and Mississippi
       Rivers. Final Report.  Missouri Dept.  Conservation, Columbia.

Carlson, D. M., W. L. Pflieger, L.  Trail and P.  S.  Haverland. 1985.  Distribution, biology
       and hybridization of Scaphirhynchus albus and S. platorynchus  in  the Missouri and
       Mississippi rivers.  Environ. Biol. Fishes  14:51-59.

Christiansen,  C. C., L W. Hesse and  B. Littell.  1991.  Contamination of the channel
       catfish (Ictalurus punctatus) by organochloride pesticides and polychlorinated
       biphenyls in the Missouri River.  Trans. Nebraska Acad. Sci. 18:93-98.


                                         231

-------
Clapp, J. R.  1977. Wildlife habitat evaluation of the unchannelized Missouri in South
      Dakota.  Master's thesis, South Dakota State Univ., Brookings.

Cuffney, T. F. 1988.  Input, movement and exchange of organic matter within a subtropical
      coastal blackwater river-floodplain system.  Freshwat. Biol. 19:305-320.

Davies, B. R. and K. F. Walker. 1986. River systems as ecological units.  An introduction
      to the ecology of river systems. Pages 1-23 in  B. R. Davies and K. F. Walker (eds.),
      The ecology of river systems.  Dr. W. Junk, The Hague, Netherlands.

Deacon, J. E., G. Kobetich, J.  D. Williams and S. Contreras.  1979. Fishes of North
      America  endangered, threatened or of special  concern: 1979. Fisheries 4(1):29-44.

Dodge, D. P. (ed).  1989.  Proc. international large river symposium (LARS).  Can. Spec.
      Publ. Fish. Aquat. Sci 106, Dept.  Fisheries and Oceans, Ottawa.

Elwood, J. W., J. D. Newbold,  R. V. O'Neill and W. Van Winkle.  1983. Resource spiraling:
      an operational paradigm for analyzing  lotic ecosystems.  Pages 3-27 jn T. D. Fontaine
      III and S. M. Bartell (eds), Dynamics of lotic ecosystems. Ann Arbor Science, Ann
      Arbor, Michigan.

Farrell, J. R. and M. A. Tesar.  1982.  Perphytic algae in the channelized Missouri River
      with special emphasis on apparent optimal temperatures.  Pages 85-124 m L. W.
      Hesse, G. L Hergenrader,  H. S. Lewis, S. D. Reetz and A. B. Schlesinger (eds.), The
      middle Missouri River.  The Missouri River Study Group,  Norfolk, Nebraska.

Ford, J. C.  1982.  Water quality of the lower  Missouri River, Gavins Point  Dam to mouth.
      Missouri  Dept. Natural  Resources, Jefferson City.

Fremling, C.  R.,  J. L. Rasmussen,  R. E. Sparks, S. P. Cobb, C. F. Bryan and T. 0. Claplin.
      1989.  Mississippi River fisheries: a case history.  Pages  309-351 in D. P. Dodge (ed.),
      Proc. international large river symposium.  Can. Spec. Publ. Fish. Aquat. Sci. 106,
      Dept. Fisheries and Oceans, Ottawa.

Funk, J. L. and J. W.  Robinson. 1974. Changes in the channel of the lower Missouri
      River and effects on fish and wildlife. Aquatic Series No.  11.  Missouri Dept.
      Conservation, Jefferson City.

General Accounting Office. 1992,  Corp's management of ongoing drought  in the
      Missouri  River Basin. Report No. GAO/RDED-92-4. Resources,  Community and
      Economic Development Division, U.S. General Accounting  Office, Washington, DC.

Grace, T. B.  1985. The status and distribution of commercial and forage  fish  in the
      Missouri  River and their utilization of selected habitats. Evaluation of sand island
      habitat in the Missouri River. Final Report.  Missouri Dept.  Conservation, Columbia.

Grace, T. B. and W. L. Pflieger. 1989. Common and scientific names of fishes of the
      Missouri  River, Missouri. Unpublished Report.  Missouri  Dept. Conservation,
      Columbia.


                                         232

-------
Groen, C. L and J. C. Schmulbach.  1978. The sport fishery of the unchannelized and
       channelized middle Missouri River. Trans. Am. Fish. Soc. 107:412-418.

Grubaugh, J. W. and R. V. Anderson. 1989.  Upper Mississippi River: seasonal and
       floodplain forest influences on organic matter transport. Hydrobiologia 174:235-244.

Herricks, E. E. and L. L Osborne.  1985.  Water quality restoration and protection in
       streams and rivers. Pages 1-20 in J. A. Gore (ed), The restoration of rivers and
       streams-theories and experience.  Butterworth Publishers, Boston.

Hesse, L W.  1987.  Taming the wild Missouri River: what it has cost?  Fisheries 12(2) :2-9.

Hesse, L. W.  and G. E. Mestl.  1992. An  alternative hydrological cycle for the Missouri
       River based on the pre-control condition. N. Am. J. Fish. Manage, (in press).

Hesse, L. W., R. L. Brown and J. F. Heisinger. 1975. Mercury contamination from a
       polluted watershed.  J. Wildl.  Manage. 39:299-304.

Hesse, L. W., A. B. Schlesinger, G. H. Hergenrader, S. D. Reetz and H. S. Lewis. 1982.
       The Missouri River study-ecological perspective. Pages 287-300 in  L.W.  Hesse, G. L.
       Hergenrader, H. S. Lewis, S. D. Reetz and A. B.  Schlesinger (eds.), The middle
       Missouri River. The Missouri  River Study Group, Norfolk, Nebraska.

Hesse, L. W., C. W. Wolfe and N. K. Cole. 1988. Some aspects of energy flow in the
       Missouri River ecosystem and a rationale for recovery. Pages 13-29 jn N. G. Benson
       (ed), The Missouri River: the resources, their uses and values. North Central Div., Am.
       Fish. Soc., Spec. Pub. No. 8.

Hesse, L W., J. C. Schmulbach, J. M. Carr, K. D. Keenlyne, D. G. Unkenholz, J. W.
       Robinson and G. E. Mestl.  1989a. Missouri River fishery resources in relation to past,
       present and future stresses. Pages 352-371 in D. P. Dodge (ed), Proc. international
       large river symposium.  Can. Spec. Publ. Fish. Aquat. Sci. 106, Dept.  Fisheries and
       Oceans, Ottawa.

Hesse, L. W., G. R. Chaffin and J. Brabander. 1989b. Missouri River mitigation: a system
       approach. Fisheries 14(1 ):11-15.

Hesse, L. W., G. E. Mestl and J. W. Robinson. 1992. Historical trends in fish density in
       the Missouri River at the species and community level with a discussion of cause and
       effect factors.  Paper presented at 1992 Annual Meeting, Am. Fish. Soc., Rapid  City,
       South Dakota.

House Report No. 475. 1944. 78th Congress, 2nd Sess. at 26.

Hynes, H. B. N.  1989. Keynote address.  Pages 5-10 in D. P. Dodge (ed), Proc.
       international large river symposium. Can. Spec. Publ. Fish. Aquat. Sci. 106, Dept.
       Fisheries and Oceans, Ottawa.
                                         233

-------
Junk, W. J., P. B. Bayley and R. E. Sparks.  1989.  The flood pulse concept in
       river-floodplain systems. Pages 110-127 in D. P. Dodge (ed), Proc. international large
       river symposium. Can. Spec. Publ. Fish. Aquat. Sci. 106, Dept. Fisheries and Oceans,
       Ottawa.

Kallemeyn, L 1 983. Status of the pallid sturgeon, Scaphirhynchus albus.  Fisheries
Keenlyne K. D.  1 988.  Economic development from Lewis and Clark to Pick-Sloan.
       Pages 31-37 jn N. G. Benson (ed), The Missouri River: the resources, their uses and
       values. North Central Div., Am. Fish Soc., Spec. Pub. No. 8.

Lieth, H. and R. H. Whittaker (eds).  1975.  Primary productivity of the biosphere.
       Springer-Verlag, New York.

Lillehammer, A. and S. J. Saltveit (eds). 1984. Regulated rivers. Universitetsforlaget AS,
       Oslo.

Lord, W. D., S. K. Tublesing and C. Althen.  1975.  Fish and wildlife implications of upper
       Missouri River Basin water allocation. Institute of Behavioral Science, Monography 22.
       University of Colorado, Boulder.

Mestl, G. E. and L. W. Hesse.  1 992. Secondary production of aquatic insects in a
       backwater of the Missouri River. Nebraska Game and Parks Commission, Norfolk.

Minshall, G. W., K. W. Cummins, R. C. Petersen, C. E. Gushing, D. A. Bruns, J. R. Sedell
       and R. L. Vannote.  1985. Developments in stream ecosystem theory.  Can.  J. Fish.
       Aquat. Sci. 42:1045-1055.

Missouri Department of Conservation.  1991 . Rare and endangered species of Missouri
       checklist. Natural History Section, Missouri Dept. Conservation, Jefferson City.

Modde, T. C. and J. C. Schmulbach. 1973. Seasonal changes in the drift and benthic
       macroinvertebrates in the unchannelized Missouri River in South Dakota.  Proc. South
       Dakota Acad. Sci. 52:118-126.

Morris, L A., R. N. Langemeier, T. R. Russell and A. Witt Jr. 1968. Effects of mainstem
       impoundments and  channelization upon the Missouri River, Nebraska.  Trans. Am.
       Fish. Soc. 97:380-388.

Moyle,  P. B. and J. J. Cech Jr.  1988.  Fishes-an introduction to ichthyology. Prentice
       Hall, Englewood Cliffs, New Jersey.

Mulholland, P. J. 1981.  Organic carbon flow in a swamp-stream ecosystem. Ecol.
       Monogr. 51 :307-322.

National Research Council. 1992.  Restoration of aquatic ecosystems-science, technology
       and public  policy. National  Academy Press, Washington, DC.
                                        234

-------
Newbold, J. D., J. W. Elwood, R. V. O'Neill and W. Van Winkle. 1981.  Measuring nutrient
       spiralling in streams.  Can. J. Fish. Aquat. Sci. 38:860-863.

Nord, A.  and J. C. Schmulbach. 1973. A comparison of macroinvertebrate aufwuchs in
       the unstabilized and stabilized Missouri River.  Proc. South Dakota Acad. Sci.
       52:127-139.

Pallid Sturgeon Recovery Team.  1992.  Pallid sturgeon (Scaphirhvnchus albus) recovery
       plan. Region 6, U.S. Fish and Wildlife Service, Denver, Colorado.

Pavlov, D. S. and B. Ya Vilenkin.  1989.  Present state of the environment, biota and
       fisheries of the Volga River. Pages 504-514 ]n D. P. Dodge (ed), Proc. international
       large river symposium.  Can. Spec. Publ. Fish. Aquat. Sci. 106,  Dept. Fisheries and
       Oceans, Ottawa.

Petts, G.  E.  1984.  Impounded rivers. Perspectives for ecological management. John
       Wiley & Sons, Chichester.

Pflieger, W. L  and T. B. Grace. 1987. Changes in the fish fauna of the lower Missouri
       River, 1940-1983.  Pages 166-177 jn W. J. Matthews and D. C. Heins (eds),
       Community and evolutionary ecology of North American stream fishes.  Univ.
       Oklahoma Press, Norman.

Platner, W. S.  1946.  Water quality studies of the Mississippi River. U.S. Fish and Wildlife
       Service Spec. Sci. Rep. 30:1-77.

Poff, N. L 1992. Why disturbances can  be predictable: a perspective  on the definition of
       disturbance in streams. J.  N. Am. Benthol. Soc. 11:86-92.

Rasmussen, J. L. 1991. Interjurisdictional fisheries initiative for the Mississippi River
       drainage  basin-comprehensive strategic plan.  Mississippi Interstate Cooperative
       Resource Agreement, Columbia, Missouri.

Reetz, S. D.  1982.  Phytoplankton studies in the Missouri River at Ft. Calhoun and
       Cooper Nuclear Stations.  Pages 71 -83 in L. W. Hesse, G. L. Hergenrader, H. S. Lewis,
       S. D. Reetz and A. B. Schlesinger (eds), The middle Missouri River. The Missouri River
       Study Group,  Norfolk, Nebraska.

Reiley, P. W. and W. C. Johnson.  1982.  The effects of altered hydrologic regime on tree
       growth  along the Missouri River in North Dakota.  Can. J. Bot. 60:2410-2423.

Repsys, A. J. and G.  D. Rogers. 1982. Zooplankton studies in the channelized Missouri
       River. Pages  125-146 in L W. Hesse,  G. L. Hergenrader, H. S. Lewis, S. D. Reetz and
       A. B. Schlesinger (eds), The middle Missouri River.  The Missouri River Study Group,
       Norfolk, Nebraska.

Resh, V.  H., A. V. Brown, A. P. Covich, M. E. Gurtz, H. W.  Li, G. W. Minshall, S. R. Reice,
       A. L. Sheldon, J. B. Wallace and R. Wissmar. 1988.  The role of disturbance  in stream
       ecology. J. N. Am. Benthol. Soc. 7:433-455.


                                        235

-------
Risotto, S. P. and R. E. Turner. 1985. Annual fluctuation in abundance of the commercial
       fisheries of the Mississippi River and tributaries.  N. Am. J. Fish. Manage. 5:557-574.

Robinson, J. W.  1989. Missouri's commercial fishery harvest, 1987.  Missouri Dept.
       Conservation, Columbia.

Robinson, J. W.  1992. Missouri's commercial fishery harvest, 1990.  Missouri Dept.
       Conservation, Columbia.

Robison, H. W.  1986. Zoographic implications of the Mississippi River Basin. Pages
       267-285 in C.  H. Hocutt and  E. O. Wiley (eds), The zoogeography of North American
       freshwater fishes. John Wiley & Sons, New York.

Schlosser, I. J.  1991. Stream fish ecology: a landscape perspective.  BioScience
       41:704-712.

Schmudde, T. H.  1963.  Some aspects of the lower Missouri River floodplain. Ann.
       Assoc. Am. Geogr. 53:60-73.

Schmulbach, J. C., G. Gould and C. L Groen.  1975.  Relative abundance and distribution
       of fishes in the Missouri River, Gavins Point Dam to Rulo,  Nebraska.  Proc. South
       Dakota Acad.  Sci. 54:194-222.

Schmulbach, J. C., L. W. Hesse and J. E. Bush. 1992. The Missouri River-Great Plains
       thread of life.  Pages 137-158 in C. D. Becker and D. A. Nietzel (eds), Water  quality in
       North American river systems.  Batelle Press, Columbus, Ohio.

Sedell, J. R., J. E. Richey and F. J. Swanson.  1989.  The river continuum concept: a basis
       for the expected ecosystem behavior of very large rivers?  Pages 49-5 in  D. P. Dodge
       (ed),  Proc. international large river symposium.  Can. Spec. Publ. Fish. Aquat. Sci. 106,
       Dept. Fisheries and Oceans,  Ottawa.

Singh, K. P. and A. Durgunoglu. 1991. Remedies for sediment buildup. Hydro Review
       10:90-93.

Slizeski, J. J., J. L. Anderson and W. G. Dorough. 1982.  Hydrologic setting, system
       operation,  present and future stresses.  Pages 15-37 in L. W. Hesse, G. L.
       Hergenrader, H. S. Lewis, S.  D. Reetz and A. B. Schlesinger (eds), The middle
       Missouri River. The Missouri River Study Group, Norfolk,  Nebraska.

Sowards, C. L. and S. Maxwell. 1985. Missouri River bibliography.  SD-IR-86-02. U.S.
       Fish and Wildlife Service, Pierre, South Dakota.

Sparks, R. E., P. B. Bayley, S. L Kohler and L. L. Osborne.  1990. Disturbance and
       recovery of large floodplain rivers. Environ. Manage.  14:699-709.
                                         236

-------
Stalnaker, C. B.( R. T. Milhous and K. D. Bovee. 1989.  Hydrology and hydraulics applied
       to fishery management in large rivers.  Pages 13-30 in D. P. Dodge (ed), Proc.
       international large river symposium.  Can. Spec. Publ. Fish. Aquat. Sci. 106, Dept.
       Fisheries and Oceans, Ottawa.

Statzner, B. and B. Higler. 1985.  Questions and comments on the River Continuum
       Concept. Can. J. Fish. Aquat. Sci. 42:1038-1044.

Sveum, D. J. 1988. Regulation of Missouri mainstem reservoirs.  Pages 51-66 in N. G.
       Benson (ed), The Missouri River: the resources, their uses and values. North Central
       Div., Am. Fish. Soc., Spec. Pub. No. 8.

U.S. Army Corps of Engineers. 1985. Missouri River mainstem reservoirs. Summary
       of actual 1984-85 operations and operating plan for 1985-86.  USACE, Omaha,
       Nebraska.

U.S. Army Corps of Engineers. 1992. Missouri River bank stabilization and navigation
       fish and wildlife mitigation project. 1992 annual implementation report. USAGE,
       Missouri River Division.

Vannote, R.  L,  G. W. Minshall, K. W. Cummins, J. R. Sedell and C. E. Gushing. 1980.
       The river continuum concept. Can. J. Fish. Aquat. Sci. 37:130-137.

Ward, J. V. 1989. Riverine-wetland interactions. Pages 385-399 m R. Sharitz and J.
       Gibbons (eds.), Freshwater wetlands and wildlife. CONF-8603101.  DOE Symposium
       Series No. 61,  U.S. Dept. of Energy, Office of Scientific and Technical Information,
       Oak Ridge, Tennessee.

Ward, J. V. and J. A. Stanford.  1979. Ecological factors controlling stream
       zoobenthos with  emphasis on thermal modification of regulated streams.  Pages 35-56
       in J. V. Ward and J. A. Stanford, The ecology of regulated streams. Plenum Press,
       New York.

Ward, J. V. and J. A. Stanford. 1983. The serial discontinuity concept of lotic
       ecosystems. Pages 29-42 in T. D. Fontaine III and S. M. Bartell (eds), Dynamics of
       lotic ecosystems.  Ann Arbor Science, Ann Arbor, Michigan.

Ward, J. V. and J. A. Stanford.  1989.  Riverine ecosystems: the influence of man on
       catchment dynamics and fish ecology.  Pages 56-64 in D. P. Dodge (ed), Proc.
       international large river symposium.  Can. Spec. Publ. Fish. Aquat. Sci. 106, Dept.
       Fisheries and Oceans, Ottawa.

Weaver, J. E. 1960. Floodplain vegetation  of the central Missouri valley and  contacts
       of woodland with prairie.  Ecol. Monogr. 30:37-64.

Webster J. R. and B. C. Patten. 1979. Effects of watershed perturbation on stream
       potassium and calcium dynamics. Ecol. Monogr. 49:51-72.
                                        237

-------
Weithman, A. S. and G. G. Fleener.  1988.  Recreational use along the Missouri River
       in Missouri. Pages 67-78 in N. G. Benson (ed), The Missouri River: the resources,
       their uses and values.  North Central Div., Am. Fish Soc., Spec. Pub. No. 8.

Welcomme, R. L.  1985.  River fisheries. FAO Fish. Tech.  Pap. 262.

Welcomme, R. L.,  R. A. Ryder and J. A. Sedell.  1989. Dynamics of fish assemblages
       in river systems-a synthesis.  Pages 569-57 in D. P. Dodge (ed), Proc. international
       large river symposium.  Can. Spec.  Publ. Fish. Aquat. Sci. 106, Dept. Fisheries and
       Oceans, Ottawa.

Whitley, J. R. and  R. S. Campbell. 1974. Some aspects of water quality and biology
       of the Missouri River. Trans. Missouri Acad. Sci. 7-8:60-72.

Whitmore, S.  B. and K. D. Keenlyne.  1990.  Rare, threatened and endangered endemic
       species of  the Missouri River floodplain.  Report MRC-90-1.  U.S. Fish and Wildlife
       Service,  Missouri River  Coordinator's Office, Pierre, South  Dakota.
                                        238

-------
                     HYDROLOGIC MODIFICATION TO IMPROVE HABITAT
                       IN RIVERINE LAKES: MANAGEMENT OBJECTIVES.
                    EXPERIMENTAL APPROACH. AND INITIAL CONDITIONS
                                             by

                      Barry L Johnson1, John W. Barko2'4, Yuri Gerasimov3,
                     William F. James4, Alexander Litvinov3, Teresa J. Naimo1,
                            James G. Wiener1, Robert F. Gaugush2,
                             James T. Rogala2, and Sara J. Rogers2
                                         ABSTRACT

       The Finger Lakes habitat-rehabilitation project is intended to improve
physical and chemical conditions for fish in six connected backwater lakes in Navigation Pool 5 of the
upper Missouri River. The primary management objective is to improve water temperature, dissolved
oxygen concentration, and current velocity during winter for  bluegills, Lepomis macrochirus, and black
crappies, Pomoxis nigromaculatus, two of the primary sport  fishes in the lakes. The lakes will be
hydrologically altered by installing culverts to introduce controlled flows of oxygenated water into four
lakes, and an existing unregulated culvert on a fifth lake will be equipped with a control gate to regulate
inflow. These habitat modifications constitute a manipulative field experiment that will compare pre-
project (1991  to summer 1993) and post-project (fall 1993 to 1996) conditions in the lakes, including
hydrology, chemistry, rooted vegetation, and fish and macroinvertebrate communities. Initial data
indicate that the Finger  Lakes differ in water chemistry, hydrology, and macrophyte abundance.
Macroinvertebrate communities also differed among lakes; species diversity was highest in lakes with
dense aquatic macrophytes.  The system seems to support  a single fish community, although some
species concentrated in individual lakes at different times. The introduction of similar flows into five of
the lakes will  probably reduce the existing physical and chemical differences among lakes.  However,
our ability to predict the effects of hydrologic modification on fish populations is limited by uncertainties
concerning both the interactions of temperature, oxygen, and current in winter and the biological
responses of  primary and secondary producers.  Results from this study should provide guidance for
similar habitat-rehabilitation projects in large rivers.
National Fisheries Research Center, U.S. Fish and Wildlife Service, P.O. Box    818, La Crosse, Wl
54602, USA

Environmental Management and Technical Center, 575 Lester Drive, Onalaska, Wl    54650, USA

Institute of Biology of Inland Waters, Russian Academy of Sciences, Borok,    Nekuzkyi Raion,
Yaroslavskaya Obi. 152742, Russia

4Eau Galle Limnological Laboratory, Waterways Experiment Station, U.S. Army      Corps of
Engineers, P.O. Box 237, Spring Valley, Wl 54767, USA
                                            239

-------
                                   INTRODUCTION

    Large rivers are diverse, dynamic ecosystems that are heavily affected by human
activities (Hynes 1989).  Yet information on the structure and function of large riverine
ecosystems, needed to establish effective management and rehabilitation policies, is scant
(Regier et al. 1989; Ward and Stanford 1989). In addition, few of the potential management
approaches for large rivers have been critically evaluated (Petts et al. 1989).

    The Mississippi River is the largest river in North America and one of the world's major
rivers in size, physical diversity, and biotic productivity (Fremling and Claflin 1984; Fremling
et al.  1989). The upper Mississippi River extends 1078 km from St. Anthony Falls,
Minnesota, to the confluence with the Ohio River.  In the 1930s, the upper Mississippi River
was divided by a series of locks and dams into 29 navigation pools to enhance commercial
navigation (a navigation  pool is the reach between two consecutive locks and dams). The
upstream portions of many navigation pools contain multiple channels characteristic of the
unimpounded river, whereas the downstream  reaches are similar to reservoirs.  This
impounded river contains large areas of diverse habitats that support economically
important sport fisheries (Fremling et al. 1989).

    Man's activities have adversely affected the quality of the riverine environment. Much of
the watershed of the upper Mississippi is intensively cultivated, and many tributary streams
deliver substantial loads of nutrients, pesticides, and sediment from eroding farmland. This
sediment accumulates in quiescent riverine lakes and backwaters and in Impounded areas
behind dams (McHenry et al. 1984) - all  important habitats for fish.  Substantial loadings of
point-source pollutants are discharged into the river from bordering Industrialized
metropolitan areas.  For example, organic pollution depleted dissolved oxygen in 100 km of
river below metropolitan Minneapolis-St. Paul, Minnesota, for several decades (Fremling
1964;  Fremling and Johnson 1990). Water quality in this reach improved since the late
1970s, because of better treatment of wastewater in the metropolitan area. Yet the river is
still a  nutrient-enriched, eutrophic system, where fish are occasionally affected by oxygen
depletion in both winter and summer (Bodensteiner and Lewis 1992).

    Federal and state resource-management agencies have initiated several rehabilitation
and enhancement projects to improve certain degraded riverine habitats for fish and wildlife
as one component of the Environmental Management Program for the Upper Mississippi
River System.  In this paper, we describe the Finger Lakes  Project, a habitat-rehabilitation
project intended to improve water quality for fish in a series of backwater lakes.  We present
the management objectives of the project and outline the experimental approach used to
evaluate the limnologic, population, and community responses to hydrologic modification of
the lakes.  Last, we summarize baseline information on the initial physical and chemical
conditions and biotic communities (rooted vegetation, fish,  and benthic macrolnvertebrates)
in the lakes during summer and fall 1991.
                            MANAGEMENT OBJECTIVES

    The Finger Lakes system near Kellogg, Minnesota, contains six connected Jakes (Gear,
Lower Peterson, Schmokers, Third, Second, and First) in Navigation Pool 5 of the upper
Mississippi River, just below the dike at Lock and Dam 4 (Figure 1). Some of these shallow,
                                         240

-------
        PETERSON LAKE (POOL 4)
                                      SPILLWAY
                        LOWER
                       PETERSON
                         LAKE
                                                                 OUTLET CHANNEL
 Location of Study Area
FIGURE 1.  Map of the Finger Lakes study area in Pool 5 of the upper Mississippi River near
           Kellogg, Minnesota, USA (star on inset map), at an elevation of 201.2 m above
           mean sea level, showing depth contours (0.5-m intervals), water sampling
           stations (•), and distribution of aquatic macrophytes (cross-hatched areas)
           during summer 1991.

                                      241

-------
eutrophic lakes have periodically undergone oxygen depletion In winter and summer
(Grunwald 1977), which reduces their suitability for fish. A culvert through the dike allows
flow into Lower Peterson Lake from Upper Peterson Lake, which Is part of the impoundment
behind Lock and Dam 4.  A mean flow of 2.3 m3/s through the culvert provides sufficient
dissolved oxygen to Lower Peterson and Schmokers lakes for fish, but reduces water
temperatures in winter to near 0°C.  Many sport fishes in the upper Mississippi River are
intolerant of such low temperatures, particularly in the presence of flow (Sheehan et al.
1990). At normal pool elevation, all six lakes drain through a common outlet to the main
channel of the Mississippi River (Figure 1).

    The primary management objective of the Finger Lakes Project is to provide suitable
water temperature, dissolved oxygen concentration,  and current velocity in winter for
bluegills and black crappies the two primary sport fishes in the lakes. To achieve these
objectives, the culvert on Lower Peterson Lake will be equipped with a control gate to
regulate flow, and culverts will be installed to introduce and regulate flow into Clear, First,
Second, and Third lakes.

    The principal objectives of our research are to determine (1) the response of the Finger
Lakes ecosystem to these  hydrologic modifications,  and (2) the flow rates that will optimize
winter temperature, dissolved  oxygen concentration, and current velocity in these lakes for
bluegills and black crappies.   We consider the habitat modifications to be a manipulative
field experiment that will entail comparisons of the pre-project and post-project status of lake
hydrology, lake chemistry,  rooted vegetation, fishes, and benthic macroinvertebrates.  The
study began in 1991 and will continue through 1996.  Results from this study should provide
guidance for other habitat-rehabilitation projects on the river.
                            EXPERIMENTAL APPROACH

HYDROLOGIC MODIFICATION

    Three new culverts will be installed through the dike of Lock and Dam 4 in 1992 and will
become operational by summer 1993. Two culverts, each 0.9 m in diameter, will provide
flow to Clear and Third lakes.  A third culvert, 1.2 m in diameter, will provide flow to both
Second and Rrst lakes through a junction box.  Gates for controlling flow will be installed on
the Inlets of each new culvert, on the existing culvert in  Lower Peterson Lake, and on the
junction box.

    Operation of the culverts and determination of their flow rates will be based on the
oxygen deficit of each lake. Preliminary estimates, based on lake morphometries and
limited chemical data, indicated that flow rates averaging about 0.8 m3/s for Lower Peterson
Lake and 0.2 to 0.3 m3/s for other lakes should maintain dissolved oxygen concentrations of
at least 5 mg/L in winter (U.S. Army  Corps of Engineers, St. Paul, Minnesota, unpublished
data).  Total inflow through all culverts would average about 1.7 m3/s for the entire system,
which is less than the 2.3 m3/s that now flows into Lower Peterson Lake.
                                         242

-------
 PHYS1COCHEMICAL

     Annual water budgets will be modeled with the following mass-balance equation:

  (1)  culvert   +  groundwater  +    dike     =   channel   +    evaporation.
        Inflow           inflow           seepage       outflow

     Inflow and outflow will be monitored every 2 weeks from August 1991 through
 September 1996.  Evaporative losses will be estimated from weather data. The sum of
 groundwater Inflow and seepage through the dike will be estimated by difference.

     We will determine flow patterns and water exchange rates within the system  from the
 movement of a fluorescent dye injected into the water.  Dye studies will  be conducted
 before the installation of the culverts (winter 1991-1992) and again  during culvert  operation
 (winter 1994-1995) under conditions of similar river discharge and water elevation.

     We will sample water at two stations in each lake (Figure 1), at weekly intervals during
 ice-free periods, and less frequently during ice cover. At each station, Secchi disk
 transparency will be measured, and depth profiles of temperature,  dissolved oxygen,  pH,
 and conductivity will be determined jnjjtu. Water samples collected at depths of 0.1  m
 below the surface and 0.1 m above the sediment-water interface at each station will be
 analyzed for  alkalinity, seston dry mass, paniculate organic matter, dissolved organic
 carbon, total  and dissolved nitrogen, ammonia-nitrogen,  total and dissolved phosphorus,
 soluble-reactive phosphorus, dissolved iron, dissolved manganese, chlorophyll, and
 phaeopigments.


 VEGETATION

     We will survey and map aquatic vegetation annually with a combination of transect
 sampling, visual surveys,  and  aerial photography.  Transects will be established at 50-m
 intervals across each lake. At 10-m intervals along each transect, plants will be sampled
 with a plant rake to estimate the relative density of plant  species within a 1-m radius.  In the
 area between transects, vegetation beds larger than 5 m in diameter will be visually
 surveyed to determine species composition and bed size. Aerial photographs (scale
 1:15,000) taken In August of each year will be used to map the extent of vegetation beds
 identified In field surveys.


 MACROINVERTEBRATES

     Macroinvertebrates will be sampled from Clear, Lower Peterson, and Third lakes to
 determine taxonomic composition, density, and biomass. Sampling will  be conducted twice
 annually, in April or May (coinciding with minimum macrophyte abundance) and in August
 (coinciding with maximum macrophyte abundance). In each lake, a maximum of 15 sample
 sites will be randomly selected in each of three habitats:  benthic substrate in non-vegetated
areas, benthic substrate under vegetation, and vegetation. Three samples will be collected
at each site.  Substrates will be sampled with a ponar grab (0.0225 m2) and washed through
a 595-pim  mesh sieve. Macroinvertebrates associated with vegetation will be sampled by
                                         243

-------
cutting a standard length of non-branching stem (50 cm for Nuphar. Nvmphaea. and
Nelumbo: 25 cm for other plant species) from the uppermost portion of five plants of the
predominant species at a site. Ail samples will be preserved in 10% formalin.
FISH

    We will describe the fish community of the lakes and the population biology of three
target species (blueglll, black crappie, and largemouth bass Mlcropterus salmoldes) before
and after culvert operation. Population and community characteristics will be assessed in
relation to habitat features both within lakes and among lakes.  Second Lake will not be
routinely studied because access to the lake is restricted for both fish and investigators.

    Standardized sampling will be done in each lake with fyke nets and electrofishing during
spring, summer, and fall (Nielsen and Johnson 1983) to determine community composition
and the relative abundance of fish species.  Target species will  be measured and weighed,
and scale samples for age estimation and back-calculation of growth will  be taken in spring.
Diets of each target species will be described by analyzing stomach contents from fish
captured seasonally. Age 1 and older fish of each target species will be fin-clipped for
mark-recapture population estimates.  Age 2 and older fish of each target species
(approximated by total lengths of at least 150 mm for biuegills,  170 mm for black crappies,
arid 200 mm for largemouth bass) will be tagged with individually numbered anchor tags.
Anglers will be surveyed in winter to estimate fishing effort and harvest. The survival rates of
target species will be estimated directly from spring and fall mark-recapture studies and also
from the return rate of tagged fish (Ricker 1975; Burnham et al. 1987).

    Movements of fish within the Finger Lakes will be determined in part  by the locations of
recaptured, tagged fish caught by anglers and during our sampling.  Fyke nets will also be
set in the outlet channel during spring, summer, and fall to determine rates of immigration
and emigration for the system.

    The effect of habitat modification on the movement of biuegills and black crappies will
be further examined in two radio-telemetry studies - both in winter -one  before and one
after installation of the culverts.  Radio transmitters will be implanted in 20 biuegills and  20
black crappies during each study. Fish will be located at least weekly, and dissolved
oxygen concentration, temperature, and current velocity will be measured at each fish's
location.
DATA MANAGEMENT

    We will determine spatial coordinates based on the Universal Transverse Mercator
System for all sample points and transects, and will develop Geographic Information System
(GIS) data bases for both the pre- and post-modification periods. These data bases can be
combined to determine the distribution of habitat features over Jime.  GIS maps will be
developed to examine both spatial and temporal changes in habitat and fish populations.
These spatial analyses will complement traditional  statistical analyses, and will help to
identify  spatial patterns and the relations  between  changes in habitats and biota.
                                         244

-------
                                INITIAL CONDITIONS

    Data from the first sampling period of this study were collected in late summer and fall
In 1991 and are summarized herein.
 MORPHOMETRY

    A bathymetric survey conducted in July 1991 provided morphometric information for the
 lakes (Table 1). The total surface area of the Finger Lakes system is 100 ha; individual lakes
 range in area from 8 to 19 ha. Lower Peterson Lake is generally deeper (mean depth 1.2 m;
 maximum, 4.3 m near the culvert) than the other lakes, whose mean depths ranged from 0.6
 to 0.9 m and whose maximum depths ranged from about 1 to 2 m.  The only shoreline
 development in the system consists of 25 houses and a boat ramp,  all located on the
 western shore of Clear Lake.
HYDROLOGY

    At present, the hydrology of the Finger Lakes depends primarily on the inflow through
the culvert into Lower Peterson Lake and on outflow through the outlet channel. Pool
elevations in Upper Peterson Lake are nearly constant at 203.1 m above mean sea level,
except during spring floods, and have never exceeded 206.8 m, the level needed to overflow
the dike. Therefore, the flow rate through the culvert is primarily controlled by the elevation
of the Mississippi River in the tailwater below Lock and Dam 4, which is determined by
discharge through the dam. Inflow through the culvert remains at about 2.3 m3/s at
dam-discharge rates up to 1,700 m3/s. Higher dam-discharge rates raise tailwater
elevations, reducing flow through the culvert to a minimum of about 1  m3/s.  Discharges
through Lock and Dam 4 exceed 1,700 m3/s only about 10% of the time, primarily during
spring floods; thus, flow through the culvert does not vary widely during  the year.

    During late summer and fall 1991, the volume and surface elevation  of the Finger Lakes
was closely correlated with elevation of the Lock and Dam 4 tailwaters (Figure 2A) and was
regulated primarily by flow through the outlet channel. Inflow through  the culvert and
outflow through the outlet channel were similar (Figure 2B) except in mid-September, when
increased  discharge through Lock and Dam 4  raised the tailwater; this increase reversed the
direction of flow through the outlet channel (negative value in Figure 2B), increasing the
volume of the lakes. A rapid decrease in tailwater elevation in late September produced
large outflows, rapidly decreasing lake volume.
PHYSICOCHEMICAL

    Mean physicochemical characteristics varied among lakes during late summer and fall
1991 (Table 2).  We present only data collected from 0.1 m below the surface because
values from 0.1  m below the surface and 0.1 m above the sediment-water interface were
similar within lakes, except as noted.
                                         245

-------
 TABLE 1. MORPHOLOGY OF THE FINGER LAKES, MINNESOTA, AT A WATER SURFACE
                ELEVATION OF 201.2 M ABOVE MEAN SEA LEVEL
Lake
Clear
Lower Peterson
Schmokers
Third
First
Outlet channel
(including bays)
Total system
Mean
depth
(m)
0.8
1.2
0.9
0.6
0.6
0.6

Maximum
depth
(m)
1.4
4.3
1.3
2.0
1.1
1.7

Surface
area
(ha)
10
8
19
15
10
8
100*
Volume
(1000 m3)
84
94
164
95
56
49
542
'Includes 30 ha of surface area in isolated lakes, contiguous ponds, and wetlands.
                                   246

-------
      40
  o

  x
  £   35
  0)
  E
  o
      30
                       Volume
                                              *
                       Tailwoter Elevation  /M-,
                                                                     202
                                                                            201.5
                                                                           c
                                                                           O
                                                                          "-*->
                                                                           o
                                                                          _a>
                                                                          LJ
                 August
                             September

                               Month
                        October
 0
to
 2:
 0
6

5

4
 £    0

      -1

      -2
                    Inflow
                                                  1
                 August
September

  Month
                                                    October
  FIGURE 2.  Hydrologic characteristics of the Finger Lakes, Minnesota, during August
            through October 1991.  (A) Volume of the Finger Lakes and elevation (meters
            above mean sea level) of the tailwaters of Lock and Dam 4. (B) Inflow from the
            culvert in Lower Peterson Lake and outflow through the Finger Lakes outlet
            channel.
                                      247

-------
      TABLE 2.  MEAN VALUES (RANGE IN PARENTHESES) FOR SELECTED PHYSICOCHEMICAL
CHARACTERISTICS OF THE FINGER LAKES, MINNESOTA, MEASURED OR SAMPLED AT 0.1 M BELOW THE
 WATER SURFACE. MEANS WERE CALCULATED FROM DATA FROM TWO SAMPLING LOCATIONS PER
   LAKE ON 10 SAMPLING DATES DURING AUGUST THROUGH OCTOBER, 1991 (N = 20 PER LAKE).
Variable
Temperature, °C
Dissolved oxygen,
mg/L
Conductivity,
pS/cm
PH
Alkalinity,
mg/L (as CaCO3)
Seston, mg/L
Paniculate organic
matter, mg/L
Dissolved organic
carbon, mg/L
Ammonia-nitrogen,
mg/L
Chlorophyll,
mg/m3

Gear
15.0
(6.4-24.6)
8.9
(5.5-12.4)
303
(282-341)
7.8
(7.2-8.5)
143
(133-150)
39.6
(11.7-71.9)
13.4
(9.1-21.3)
19.9
(8.0-29.2)
0.08
(0.01-0.69)
79.0
(32.1-150)

Lower
Peterson
16.3
(9.0-24.6)
9.0
(6.4-11.8)
467
(399-521)
8.0
(7.6-8.4)
173
(153-189)
12.0
(7.0-18.7)
3.1
(0.7-5.2)
25.0
(17.9-30.1)
0.05
(0.01-0.2)
13.9
(5.6-33.2)
Lake
Schmokers
16.3
(8.1-25.3)
9.3
(6.8-11.9)
467
(395-561)
8.0
(7.8-8.5)
171
(151-184)
21.8
(6.9-58.7)
4.9
(2.5-15.2)
25.1
(14.6-29.3)
0.05
(0.01-0.11)
20.0
(10.7-45.4)

Third
15.4
(7.8-24.4)
7.3
(4.4-10.6)
392
(367-452)
7.4
(7.1-7.8)
194
(177-208)
10.4
(4.6-22.2)
4.5
(1.4-13.2)
21.3
(13.0-30.9)
0.18
(0.01-0.5)
26.0
(2.1-76.2)

First
16.6
(7.9-27.1)
10.6
(7.3-17.8)
432
(366-519)
8.1
(7.7-8.9)
173
(154-181)
31.9
(8.3-64.6)
7.9
(2.7-18.9)
24.3
(11.3-30.1)
0.07
(0.01-0.41)
43.4
(14.4-114)
                                      248

-------
     Mean water temperatures were similar among lakes, which indicates that inflow to
 Lower Peterson Lake had little effect on temperature during summer and fall 1991.
 Conductivity, pH, alkalinity, and dissolved organic carbon were similar among Lower
 Peterson, Schmokers, and First lakes, but differed from values in Clear and Third lakes.
 These data suggest that flow from the culvert into Lower Peterson Lake moved directly
 through Schmokers Lake to the system outflow, with little water entering Third and  Clear
 lakes.  Conductivity, alkalinity, and temperature were lowest in Clear Lake, probably because
 of the  influence of groundwater inputs. Mean and minimum dissolved oxygen
 concentrations were lowest in Third Lake, the only lake that exhibited temperature and
 oxygen stratification during summer 1991.  Minimum dissolved oxygen concentrations
 approached zero in the hypolimnion of Third Lake, and were associated with high
 concentrations of ammonia-nitrogen.  Lower Peterson Lake, the deepest lake, did not
 stratify, probably because of mixing caused by inflow from the culvert.

     Mean concentrations of seston, paniculate organic matter, and chlorophyll were highest
 in Clear Lake, which suggests higher primary productivity than in the other lakes (Table 2).
 In contrast, concentrations of these biogenic components were generally lowest in  Lower
 Peterson Lake, but progressively higher In Schmokers and First lakes.  These data  indicate
 that phytoplankton production increases from the inflow to the outflow and suggest that
 there was a net export of biogenic matter from the system.
VEGETATION

    Eleven species of submersed and floating-leaved aquatic vegetation were found in the
Finger Lakes (Table 3). Based on our samples and on aerial photographs taken in 1991,
Third  Lake was the most heavily vegetated and dear Lake the most sparsely vegetated of
the Finger Lakes (Figure 1). Nelumbo was the most common macrophyte in Clear Lake,
whereas some combination of Ceratophvllum. Myriophyllum. and Nymphaea dominated in
the other lakes. Plant beds generally occurred in strips near shorelines. In First and Third
lakes, the upper third of each lake and most protected bays contained dense beds of
Ceratophvllum and often contained the free-floating macrophyte, duckweed (Lemna spp.).
Beds  of Mvriophyllum existed offshore in Schmokers Lake, and small beds of Nelumbo were
found in Lower Peterson, Schmokers, and Gear lakes. The other six plant species were
rare.
MACROINVERTEBRATES

    We collected 276 samples of macroinvertebrates from three of the Finger Lakes in
August 1991. Based on initial analyses of about 10% of the samples from each lake and
habitat type, the benthos were dominated by digochaetes, followed by chironomids, and in
Third Lake, by the amphipod Hyalella azteca (Table 4).  Amphipods and chironomids were
the dominant macroinvertebrates on vegetation.  Odonate larvae, primarily Enallaama  spp.,
were common on vegetation but rare elsewhere.  Bivalve molluscs were common in
sediments from  non-vegetated areas of Lower Peterson Lake.  Diversity of
macroinvertebrates was highest in Third Lake, which had 26 taxa (Table 4).  Densities  of
macroinvertebrates in vegetated areas (benthic plus plant-associated) were higher than in
                                     249

-------
TABLE 3. FREQUENCY OF OCCURRENCE (PERCENT OF SITES CONTAINING THE
  SPECIES) OF SUBMERSED AND FLOATING-LEAFED AQUATIC PLANTS IN THE
 FINGER LAKES, MINNESOTA, IN AUGUST 1991. "P" INDICATES THAT THE PLANT
     WAS PRESENT IN THE LAKE, BUT NOT SAMPLED ON TRANSECTS.
Lake
Species

Ceratophytlum
demersum
Vallisneria
americana
Myriophvllum so.
Elodea
canadensis
Najas
flexHis
Potamoqeton
americanus
Potamoqeton
zosteriformls
Potamoqeton
crispus
Heteranthera
dubia

Netumbo
lutea
Nymphaea
tuberosa
Nuphar
tuteum
Lemnaspp.
Clear

0

0

0
0

0
0

0

0

0


P

P

0
0
Lower
Peterson Schmokers
Submersed plants
23

10

20
4

0
P

4

0

0

Floating-leafed plants
11

25

P
0

16

0

38
0

P
3

0

0

3


P

9

0
0
Third

92

0

0
2

0
P

0

0

0


0

38

P
P
First

45

0

7
1

0
4

0

2

0


0

9

0
P
                              250

-------
  TABLE 4. TAXONOMIC LIST AND RELATIVE PERCENTAGE BY NUMBER OF MACROINVERTEBRATES IN
THREE HABITAT TYPES (V = ON VEGETATION, S = SUBSTRATES UNDER VEGETATION, N = SUBSTRATES
    IN NON-VEGETATED AREAS) IN CLEAR, LOWER PETERSON, AND THIRD LAKES IN AUGUST 1991.
Macroinvertebrate
taxon
Class Ollgochaeta
Class Crustacea
Order Isopoda
Caecidotea
Order Amphipoda
Hyalella azteca
Class Arachnoidea
Hygrobates
Umnesla
Neumanla
Class Bivalvia
Family Sphaeriidae
Family Unionidae
Class Insecta
Order Diptera
Family Chironomidae
Family Ceratopogonidae
Family Psychodidae
Family ChaoborkJae
Chaoborus
Order Odonata
Suborder Zygoptera
Enallagma
Order Coleoptera
Order Ephemeroptera
Caenls
Miscellaneous taxa
Number of samples
Mean number
per sample
Clear Lake
V
0


0

0

0
0
0

0
0


100
0
0

0

0
0
0

0
0
1

8
S
91


0

0

0
1
0

0
0


4
3
0

0

0
0
0

0
2
2

69
N
95


0

0

0
1
0

0
0


0
1
0

2

0
0
0

1
0
10

34
Lower Peterson
Lake
V
5


0

43

0
0
0

0
0


37
1
0

0

0
11
1

3
0
3

59
S
48


3

2

0
1
0

0
0


38
0
0

0

1
0
2

3
3
2

93
N
32


0

0

4
0
0

18
7


29
7
0

0

0
0
0

4
0
2

14
Third Lake
V
12


1

41

0
0
0

0
0


24
0
0

0

1
7
1

2
11
4

73
S
69


6

2

0
1
2

2
0


12
0
1

1

1
0
0

0
4
4

44
N
53


0

37

0
0
2

0
0


4
0
0

1

0
0
0

1
2
3

70
                                       251

-------
 non-vegetated areas.  All macroinvertebrate communities were dominated by small
 organisms, which suggests Intensive, size-selective predatlon by fish.
FISH

    The fish assemblage of the Finger Lakes (Table 5) was typical of Upper Mississippi River
backwaters. In summer, bluegills and black crappies dominated the catch In all lakes,
followed by catostomids. Catch rates for bluegllls and black crappies were highest In Lower
Peterson Lake. In fall, catch rates increased in most lakes because age-0 fish were large
enough by then to be caught in fyke nets (Table 5). Bluegills and black crappies were
concentrated In Third and First lakes. Catches hi Clear, Lower Peterson, and Schmokers
lakes were dominated by young-of-the-year white bass Morone chrvsops and gizzard shad
Dorosoma cepedianum.

    These data agree with population estimates for age-1 and older fish in fall 1991, which
showed that numbers and densities of bluegills and black crappies were greatest in Third
Lake  (Table 6). Largemouth bass, the primary plscivore in the system, are less vulnerable to
fyke nets,  but population estimates indicated that largemouth bass were concentrated in
Lower Peterson Lake (Table 6).

    Tag returns from anglers indicated that fish  moved freely within the Finger Lakes
system, and that few fish left the system. Through  31 March 1992, 49%  of the 125 bluegill
tags and 58% of the  73 black crappie tags received from anglers were from fish recaptured
in a lake other than that where they were tagged.  Only two bluegills and four black
crappies, representing only 3% of all tag returns, were recaptured outside the Finger Lakes
system. These initial data suggest that the Finger Lakes system forms a spatial  boundary
for  populations of bluegill and black crappie, but that these populations are not confined to
individual lakes within the system.

    Preliminary analysis of stomach contents from  fish sampled in fall 1991 showed that
both bluegills and black crappies fed mainly on macroinvertebrates, mostly amphipods and
dipterans.   Zooplankton were important in the diets of young-of-the-year fishes.  Adult black
crappies occasionally ate fish, primarily gizzard shad and Lepomis spp.  Largemouth bass
consumed  mostly gizzard shad. Lepomis spp.. and darters Etheostoma.  but smaller bass
also ate macroinvertebrates.
                        CONCLUSIONS AND UNCERTAINTIES

    Initial data Indicate that the Finger Lakes differ in water chemistry, hydrology, and
macrophyte abundance.  Clear Lake had the highest chlorophyll concentrations, probably
received groundwater inputs with lower conductivity, and had few aquatic macrophytes.
Lower Peterson and Schmokers lakes, which were heavily influenced by inflow from the
culvert, were both low in chlorophyll and paniculate organic matter, had high conductivity,
and had  moderate production of aquatic macrophytes. Third Lake, which was intermediate
in physicochemical characteristics, contained dense aquatic macrophytes, and exhibited
oxygen depletion during summer.  The upper portion of First Lake was similar to Third Lake,
but the lower portion, near the system outlet, was influenced by flow from the culvert and
                                         252

-------
1
Z
Si
ii
oib
feS

§i
^co"
p£
•-CD
Q Z
"• UJ
col
UJ *-
QOC


OT0

feZ
CDZ
So.
 .o
•^ z
l§
§
a
              _  c
              -
              IJ8
                1
             II
                    T  •*  *t  •+  *t O  tn  *-  U)    m
                    «docM«6c6»;     p  1;  CM  N
                   oiddcMO>«-:o*-"dd'-:
                    i-'**KO)'*«>IO*'«t«)
                    cMdoT-c6CMK
                   CMOd^COCMOCMd'-'cod
                      SCO*  CMtOlO»-tO«aoOK«O
                      d d  co  o)  •*'  d  CM r-'  d  T^  »-"
                    IO(pCDCMCO(Oi-;peOlpCMCO
                   dddcMd^di-d^dd
                   COCO<«CMr-cO'-:'^IOCON:CO
                   dddcodiodiodddr^
                   o  o  o «-
                                   p  00)  CO i-  O>  •-  p
                                ocddcMd'-:d»-:
                   '-;CM»^iocM^^r'«;i-:cO'r'*
                   ddv-cMdo>do>d>-dd
                   o>cot.  >O'-o>S;cq
                   dWdtodcodio^d^d
                      I
midae
spp.
r Lepp

oxis nia
                                            §.
Per

E
^
253
                                                            8
                                                            8
                                                                      «<>
                                                                   *"  fu
                                                                      CM
                                                                   a
                                                            a     a ?
                                                            S     S 5
                                                                     CO
                                                                  8  2
                                                            CM
                                                            CM
                                                                      IO
                                                                   8  8
                                                                     CO

                                                                  8 JM
                                                            I  §
                                                            111
                                                            III
                                                            III
                                                                     t

                                                                     I

                                                                     1
                                                                          <3
                                                                                     •c

                                                                                     'o
                                                                                     «
                                                                                     31
                                                                                     Q
                                                                                     |
                                                                                       1

                                                                                       S

                                                                                       T
                                                                                       (t

                                                                                       §

                                                                                       §
                              1   I
                                                                                       i
                                                                                      s
                                  -JL
                                  «  !
Le
Q



fl

1|

1^

-------
£8££
IsS^la
9 ""- w i 5 t
Crs-WH-Cj^
  IO

      2Z
      Of—
  DC UJ —
55_iQ  .
UJ UJ Z ,-
^8Bs£
m uj
    = LU
                   0)
                   I
               3
                   O
                   2

                   I
                   a.

                   o
8
o
&
co
I  $
*S
   TI-"
                        n
                        co
                        o>
                        en
                        CO
                           r-
                           CM,

•
                                             >
                                       254

-------
 exhibited characteristics similar to Schmokers Lake.  Macro-invertebrate communities differed
 among lakes, with aquatic macrophytes providing an important substrate for colonization.
 The macroinvertebrate community was most diverse in Third Lake, where macrophytes were
 most dense, but was composed almost exclusively of chironomids and oligochaetes in dear
 Lake, where macrophytes were sparse. The Finger Lakes system appears to have a single
 fish community, although some species concentrate in individual lakes at different times.

     The extensive data base collected during the pre-culvert studies should allow us to
 estimate optimal temperature, dissolved oxygen concentration, and current velocity, and to
 then re-estimate the inflows needed to achieve optimal conditions in each lake. Although
 the estimated inflows may change,  the flow regime after culvert installation will presumably
 involve simitar,  low rates of inflow to each lake. Lower Peterson and Schmokers lakes will
 have reduced flows relative to present conditions, whereas flows will increase in the other
 lakes.  Total inflow to the system should  be near or below present levels.

     Responses of the system to modification in the flow regime are presently difficult to
 predict, particularly for the fish populations. Fish may respond to changes in habitat, trophic
 dynamics, or species interactions through altered survival, growth, reproduction, or
 emigration.  Changes in any of these fish-population variables could affect the relative
 abundance  of species and the community composition of fishes.

    The introduction of similar, low flows into each lake will probably reduce the
 physicochemical differences among lakes that existed in 1991. Inflow may eliminate anoxia
 in summer and increase available habitat for fish and macroinvertebrates. Winter oxygen
 concentrations  can be  increased by introduced flows, improving habitat for aquatic biota,
 but this could also decrease water temperature and increase current velocity, which could
 be stressful  to certain fishes - including our target species (Sheehan et al. 1990;
 Bodensteiner and Lewis 1992). Consequently, the flow rates needed to create optimal
winter conditions for Wuegills and black crappies must be accurately determined.

    Increased flows could affect the production of macrophytes in various ways.  Physical
disturbance caused by flows may eliminate macrophytes in some areas, particularly near the
mouths of the culverts. Changes in transparency can modify the depth of the photic zone,
but it is difficult to predict how increased flow will alter transparency. Increased inflow  may
reduce transparency by increasing the concentration of suspended solids or by stimulating
phytoplankton production. Conversely, increased flow may increase transparency by
flushing phytoplankton, suspended  solids, and non-rooted aquatic plants, such as duckweed
and Ceratophyllum. from the system.  Changes in vegetation abundance may affect
spawning and nursery habitat for many fish species.

    Fish populations in the Finger Lakes may be indirectly affected by trophic changes. The
food of target fishes seems to consist primarily of macroinvertebrates.  Increased oxygen
concentrations  in winter may improve the over-winter survival of macroinvertebrates, thereby
increasing the food supply for target fishes.  Population densities of macroinvertebrates
seem strongly related to the abundance of aquatic plants; thus, changes in aquatic
macrophytes could greatly influence the food chain supporting key fishes.

    Our ability to predict the effects of hydrologic modification on fish populations is limited
by uncertainties in  the physical responses of the Finger Lakes system, particularly the
                                          255

-------
interactions of temperature, dissolved oxygen, and current velocity in winter, and by
uncertainties in the responses of primary and secondary producers. The ability to regulate
flows should allow us to reduce undesirable responses and to determine appropriate inflows
for each lake that will optimize the quality of habitat across the system as a whole.

    The response of this system to a modified hydraulic regime could not be predicted from
controlled laboratory experiments because of the size and complexity of the system.
Questions about systems of this size can best be answered when management programs
are treated as large-scale field experiments whose results are carefully evaluated (Walters
1986).  Results from this project should yield hypotheses concerning the responses of
backwater systems to modified flows which can be used to develop new management
strategies for similar ecosystems.
                                ACKNOWLEDGMENTS

     We thank the Minnesota Department of Natural Resources, especially Mike Davis and Al
 Stevens, for field assistance, manuscript review, and cooperation in all aspects of this study.
 The manuscript was also reviewed by Steve Gutreuter, William Swink, and Steve Zigler.
 Partial funding for this study was provided by the U.S. Army Corps of Engineers through the
 Environmental Management Program.
                                        256

-------
                                    REFERENCES

 Bodensteiner, L R., and W. M. Lewis.  1992.  Role of temperature, dissolved oxygen, and
     backwaters in the winter survival of freshwater drum (Aplodlnotus qrunnlens) in the
     Mississippi River. Canadian Journal of Fisheries and Aquatic Sciences 49:173-184.

 Burnham, K. P., D. R. Anderson, G. C. White, C. Brownie, and K. H. Pollock. 1987.  Design
     and analysis methods for fish survival experiments based on release-recapture.
     American Fisheries Society, Monograph 5, Bethesda, Maryland, USA. 437 pp.

 Fremling, C. R.  1964.  Mayfly distribution indicates water quality on the upper Mississippi
     River.  Science 146:1164-1166.

 Fremling, C. R., and T. 0. Qaflin.  1984.  Ecological history of the upper Mississippi River.
     In: Contaminants in the upper Mississippi River, J. G. Wiener, R. V. Anderson, and D. R.
     McConville, eds.  Butterworth Publishers, Boston, Massachusetts, USA.  pp. 5-24.

 Fremling, C. R., and D. K. Johnson.  1990.  Recurrence of Hexaaenia mayflies demonstrates
     improved water quality in Pool 2 and Lake Pepin, upper Mississippi River. In: Mayflies
     and stoneflies, Proceedings of the International Conference on Ephemeroptera, Vol 5,1.
     C. Campbell, ed. Kluwer Academic Publishers, Norwell, Massachusetts, USA.  pp. 243-
     248.

 Fremling, C. R., J. L Rasmussen,  R. E. Sparks, S. P.  Cobb, C. F. Bryan, and T. O. Claflin.
     1989.  Mississippi River fisheries: a case history.  In: Proceedings of the International
     Large River Symposium, D. P. Dodge, ed.  Canadian Special Publication of Fisheries
     and Aquatic Sciences 106, Ottawa, Ontario, Canada,  pp. 309-351.

Grunwald, G.  1977.  Mississippi River survey:  Finger Lakes-River mile 752.6. Federal Aid
     Project Report FW-1R-21, Study VI.  Minnesota Division of Fish and Wildlife, Lake City,
     Minnesota, USA.  16  pp.

Hynes, H. B. N.  1989.  Keynote address.  In: Proceedings of the International Large River
     Symposium, D. P. Dodge, ed. Canadian Special Publication of Fisheries and Aquatic
     Sciences 106, Ottawa, Ontario, Canada,  pp. 5-10.

McHenry, J. R., J. C. Ritchie, C. M. Cooper, and J. Verdon.  1984.  Recent rates of
     sedimentation in the Mississippi River.  In: Contaminants in the upper Mississippi River,
    J. G. Wiener, R. V. Anderson,  and D. R. McConville, eds.  Butterworth Publishers,
     Boston, Massachusetts, USA.  pp.  99-118.

Nielsen, L A., and D. L Johnson,  editors. 1983. Fisheries techniques.  American Fisheries
    Society, Bethesda, Maryland,  USA.

Petts, G. E., J. G. Imhof, B. A. Manny, J. F. B. Maher. and S. B. Weisberg. 1989.
    Management of fish populations in large rivers: a review of tools and approaches.  In:
    Proceedings of the International Large River Symposium, D. P. Dodge, ed. Canadian
    Special Publication of Fisheries and Aquatic Sciences 106, Ottawa, Ontario, Canada.
    pp. 578-588.
                                          257

-------
Regier, H. A., R. L Welcome, R. J. Steedman, and H. F. Henderson.  1989. Rehabilitation of
    degraded river ecosystems.  In: Proceedings of the International Large River
    Symposium, D. P. Dodge, ed.  Canadian Special Publication of Fisheries and Aquatic
    Sciences 106, Ottawa, Ontario, Canada, pp. 86-97.

Ricker, W. E.  1975. Computation and interpretation of biological statistics of fish
    populations. Bulletin of the Fisheries Research Board of Canada 191:382 p.

Sheehan, R. J., L. R. Bodensteiner, W. M. Lewis, D. E. Logsdon, and S. D. Scherck.  1990.
    Long-term survival and swimming performance of young-of-the-year river fishes at low
    temperatures: links between physiological capacity and winter habitat requirements. In:
    Restoration of midwestern stream habitat.  Rivers and Streams Technical Committee,
    North-Central Division, American Fisheries Society, Bethesda, Maryland, USA. pp. 97-
    107.

Walters, C. J.  1986. Adaptive management of renewable resources. Macmillan Publishing
    Company, New York, New York, USA. 374 pp.

Ward, J. V., and J. A. Stanford.  1989.  Riverine ecosystems:  the influence of man on
    catchment dynamics and fish ecology.  In: Proceedings of the International Large River
    Symposium, D. P. Dodge, ed.  Canadian Special Publication of Fisheries and Aquatic
    Sciences 106, Ottawa, Ontario, Canada, pp. 56-64.
                                                 * U S GOVERNMENT PRINTING OFFICE:  1996 - 750-001/41048
                                         258

-------
m
55
o
o
O
(D
O
-~
o
"• GO
•a 5'
3- CD
< en
»_ en
CD
       CO
       CO
            o ID z m c
            5' CD ft)  3 3
            X c/5 =r. < s:
            =4- CD O  =;• CD
            ^5 fu D  O Q.

            " a s  = w
                    £.3-
             ill
               03
               ^S
               ro 3.
CD  »

2-  CD
JD  CO


TD

O

CD
£

5'
^
                            CQ
                            CD
                            O
                    «gg

                    Io|

                    m 8 a

                    ^5 I

                    85^
                    3 w o
                    CD O -,
                    ^ O CD
                      I; CD
                      °%
                      3 M
                      CD o
                      0) I

                      S3
                      8 *
                      w X

                      s3
                    lal
                    A- 0) 01
                    ni O CO
                    g =TCD
                    Q. o 3


                    8§£
                    3-8 ">
                    CD •< OJ.


                    " 5> 3

                      ^8
                      ^ CD
                      CD W

                      ?S
                      CD
                • m

                D
                      »f

-------