vvEPA
          United State*.
          ,-rvironmenfp; Protection
          Agency
                     EPA-600'3 80-078
                     July 1980
          Research and Development
Ecological Studies of
Fish Near a Coal-
Fired Generating
Station and Related
Laboratory Studies

Wisconsin Power
Plant Impact Study

-------
                RESEARCH REPORTING SERIES

Research reports of the Office of Research and Development, U S Environmental
Protection Agency, have been grouped into nine series  These nine broad cate-
gories were established to facilitate further development and  application of en-
vironmental technology  Elimination of traditional grouping  was  consciously
planned to fosler technology transfer and a maximum interface in related fields
The nine series are

      1   Environmental  Health Effects Research
      2   Environmental  Protection Technology
      3   Ecological Research
      4   Environmental  Monitoring
      5   Socioeconomic Environmental Studies
      6   Scientific and Technical Assessment Reports (STAR)
      7   Interagency Energy-Environment Research and  Development
      8   "Special" Reports
      9   Miscellaneous Reports

This report has been assigned to the ECOLOGICAL RESEARCH series This series
describes research on the effects of pollution on humans, plant and animal spe-
cies, and materials Problems are assessed  for their long- and short-term influ-
ences Investigations include formation, transport, and pathway studies to deter-
mine the fate of pollutants and their effects This work provides the technical basis
for setting standards to minimize undesirable changes in living organisms in the
aquatic, terrestrial, and atmospheric environments
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                               EPA-600/3-80-078
                                               July 1980
   ECOLOGICAL STUDIES OF FISH NEAR A COAL-FIRED
GENERATING STATION AND RELATED LABORATORY STUDIES
        Wisconsin Power  Plant  Impact Study


                        by
                 John  J.  Magnuson
                  Frank  J.  Rahel
                Michael J. Talbot
                  Anne M.  Forbes
               Patricia A. Medvick
       Institute for Environmental Studies
         University of Wisconsin-Madison
            Madison,  Wisconsin 53706
                 Grant  No.  R803971
                 Project Officer
                  Gary E. Glass
     Environmental  Research  Laboratory-Duluth
                Duluth, Minnesota
  This study was  conducted  in  cooperation with

       Wisconsin  Power  and  Light  Company,
        Madison Gas and Electric Company,
      Wisconsin Public  Service Corporation,
      Wisconsin Public  Service  Commission,
  and Wisconsin Department of Natural Resources
    ENVIRONMENTAL  RESEARCH LABORATORY-DULUTH
       OFFICE  OF RESEARCH AND DEVELOPMENT
      U.S. ENVIRONMENTAL  PROTECTION AGENCY
             DULUTH, MINNESOTA 55804
                       . V* '

                      • H*

-------
                                  DISCLAIMER
     This report has been reviewed by the Environmental Research  Laboratory-
Duluth, U.S. Environmental Protection Agency, and approved for
publication.  Approval does not signify that the contents necessarily
reflect the views and policies of the U.S. Environmental  Protection  Agency,
nor does mention of trade names on commerical products constitute
endorsement or recommendation for use.
                                      ii

-------
                                   FOREWORD
     The U.S. Environmental Protection Agency  (EPA)  was  established to
coordinate our country's efforts  toward  improving  and  defending  the quality
of the environment.  These efforts depend greatly  on research  to  monitor
environmental change and to determine health standards.

     One project the EPA is supporting through  its Environmental  Research
Laboratory in Duluth, Minnesota,  is  the  study  "The Impacts  of  Coal-Fired
Power Plants on the Environment."  The Columbia Generating  Station, a  coal-
fired power plant near Portage, Wisconsin, has  been  the  focus  of  all field
observations.  This interdisciplinary study is  conducted by the
Environmental Monitoring and Data Acquisition  Group  of the  Institute for
Environmental Studies at the University  of Wisconsin-Madison and  involves
investigators from many departments  at that same university.  Several
utilities and state agencies also are cooperating  in the study:   Wisconsin
Power and Light Company, Madison  Gas and Electric  Company,  Wisconsin Public
Service Corporation, Wisconsin Public Service  Commission, and  Wisconsin
Department of Natural Resources.

     Reports from the study will  appear  as a series  within  the EPA
Ecological Research Series.  The  topics  will treat chemical constituents,
chemical transport mechanisms, biological effects, social and  economic
effects, and integration and synthesis.

     The Columbia Generating Station has caused changes  in  nearby
wetlands.  Since the area has a diverse  fish community,  the fish-monitoring
group of the Columbia Generating  Station impact study  has been studying the
effects of habitat loss and habitat  degradation on fish.  This report
assesses the station's effects on fish reproduction  and  documents research
on the use of temperature preference to  detect  sublethal concentrations of
zinc in bluegills (Lepomis macrochirus)  and zinc selection  for tolerance
over four generations of flagfish (Jordanella floridae).
                                      Norbert A. Jaworski
                                      Director
                                      Environmental Research  Laboratory
                                      Duluth, Minnesota
                                    iii

-------
                                  ABSTRACT
     Construction of a coal-fired electric generating station on wetlands
adjacent to the Wisconsin River has permanently altered about one-half  of
the original 1,104-ha site.  Change in the remaining wetlands continues as a
result of waste heat and ashpit effluent produced by the station.  Leakage
of warm water from the 203-ha cooling lake is causing a shift in the
wetlands from shallow to deep-water marsh.  Coal-combustion byproducts  enter
the wetlands from the station's ashpit drain.  Since this area was known to
have a diverse fish community and to be a spawning ground for Wisconsin
River game fish, we studied the effects of this habitat loss and degradation
on fish populations.  In laboratory experiments we investigated the use of
temperature preference and activity as a sublethal bioassay.  In selection
experiments we examined the potential of fish to evolve metal-tolerant
populations in chronically contaminated environments.

     Three years of netting documented the continued use of this area by
spawning fish despite extensive habitat alterations.  An inventory of
potential spawning marshes along the Wisconsin River between the dams at
Wisconsin Dells and Prairie du Sac showed that the station site still
contained 22.0% of the deep-water sedge meadow and 0.8% of the shallow-water
sedge meadow likely to be used by spawning game fish.  Construction of  the
power plant resulted in a loss of 18% of the shallow water sedge meadow
formerly available in this section of the Wisconsin River.  Loss of deep
water sedge meadow was negligible.  In situ and laboratory experiments
showed that the ashpit effluent was not acutely toxic to eggs or larvae of
northern pike (Esox lucius), although some reduction in hatchability was
attributed to the flocculent precipitate found in the ashpit drain.
Analysis of population structures of northern pike showed a weak year-class
for fish hatched in the first post-operational year.  Further monitoring
will be needed to determine if the reduction was due to the generating
station or to natural factors.

     A bioassay utilizing temperature preference and activity proved no more
sensitive than bioassay methods used by previous investigators.  Bluegills
(Lepomis macrochirus), themoregulating in a temporal gradient, tended  to
increase activity and decrease preferred temperature after exposure to  2.5
mg/liter zinc. Neither change was statistically significant, however, and
both factors returned to normal levels within 2 days.

     A population of flagfish (Jordanella floridae) selected for zinc
tolerance was more resistant than, the control population for the first  two
generations but not after three generations.  The failure of continued
selection to produce ixicreasing zinc tolerance may have been caused by
inbreeding depression or by cumulative carry-over effects of zinc  passed

                                     iv

-------
from mother to offspring through  the egg  cytoplasm.   The  discrepancy between
laboratory-selection experiments  and field observations of  fish  living in
chronically metal-contaminated environments  is  discussed.

     This report was prepared with the cooperation of  faculty  and  graduate
students at the Limnology Laboratory at the  University of Wisconsin-Madison.

     Most of the funding for the  research reported here was  provided by the
U.S. Environmental Protection Agency (EPA).   Funds were also granted by the
University of Wisconsin-Madison,  Wisconsin Power and  Light  Company,  Madison
Gas and Electric Company, Wisconsin Public Service Corporation,  and
Wisconsin Public Service Commission.  This report was  submitted  in
fullfillment of Grant No. R803971 by the  Environmental Monitoring  and Data
Acquisition Group, Institute for  Environmental  Studies, University of
Wisconsin-Madison, under the partial sponsorship of  the EPA.   The  report
covers the period of July 1975-78 and work was  completed as  of April 1979.

-------
                                   CONTENTS
Foreword	
Abstract	   iv
Figures	viii
Tables  	    x
Acknowledgments	•	xii

      1.  Introduction   	  1
               Design of the fish study	5
      2.  Conclusions	••  7
      3.  Effects of the Columbia Generating  Station  on Fish Spawning . .  9
               Introduction	•	9
               Methods	10
               Results	19
               Discussion	45
      4.  Zinc Tolerance in  Four Generations  of Flagfish	52
               Introduction	52
               Literature  review of metal  tolerance 	 52
               Objective of  this study	53
               Methods	53
               Results—exposures and calculations   	 58
               Discussion	64
      5.  Use of Temperature  Preference  and Activity  as a
            Sublethal Bioassay  for the  Toxic  Effects  of Zinc to the
            Bluegill	71
               Introduction   .....  	  71
               Design of the study	  71
               Materials and methods  	  72
               Results	73
               Discussion	76

References	78

Appendices

      A.  Number of Fish Caught at Each Sampling Station,  1976-78  ...  85
      B.  Marshes Near  the Columbia Generating Station 	 ...  88
      C.  Review of Literature  on Entrainment From Cooling
            Lake Intake Structures 	  90
      D.  Review of Literature  on Acid  Precipitation	95
      E.  Review of Literature  on Alternative Disposal of  Fly Ash  . . . 101
                                     vii

-------
                                   FIGURES

Number                                                                   Page

   1  Map of study site at Columbia Generating  Station.  ...  	    2

   2  Ground-water flows before and after construction of  the
        Columbia cooling lake	    4

   3  Location of fyke nets in spawning marshes on  the Columbia  site.  .   14

   4  Fyke-net catches averaged over  the 3-yr period,  1976-78  	   20

   5  Spring water levels in the spawning grounds at  the
        Columbia site during 1976-78   	   23

   6  Catch of northern pike per unit  effort on the Columbia site
        during 1976-78  	   24

   7  Survival of northern pike eggs  hatched -in si.tu  in  the
        wetlands at the Columbia site during April  1977  ........   27

   8  Survival of northern pike larvae placed in  the  wetland
        at the Columbia site for 11 days in April  1977	29

   9  Survival of northern pike eggs  hatched in the laboratory
        during April  1978 in unfiltered water from  the ashpit
        drain, Rocky  Run Creek, and a downstream natural mixture. ...   30

  10  Survival of northern pike eggs  hatched in the laboratory
        in April 1978 using filtered  water from the asphit drain,
        Rocky Run Creek, and a downstream natural mixture  	   32

  11  ft>pulation-age  structure of northern pike caught on
        the Columbia  site in 1976,  1977, and 1978	35

  12  Ibpulation-age  structure of walleye caught  on the  Columbia
        site in spring  1977	36

  13  Major areas of  potential northern pike spawning habitat  in the
        Wisconsin River and tributaries near the
        Columbia Generating Station  	 ...   37

  14  Detailed maps  (a-g) of potential northern pike spawning
        habitats in  the Wisconsin River and tributaries  near
        the Columbia  Generating  Station  .	38-44


                                    viii

-------
15  Cumulative mortality (probit scale) as a function of exposure
      time for flagfish exposed to three zinc concentrations  ....  55

16  Procedure used in selecting for zinc resistance in
      laboratory populations of flagfish  	  56

17  Survival rates of selected and unselected flagfish populations
      over four generations	59

18  Summary of three generations of selection for zinc resistance
      in flagfish	65

19  ffedian selected temperatures of bluegill in control
      aquaria and in aquaria treated with zinc in a
      7-day experiment  .  	 ...........  74

20  Median number of tunnel passes per hour by bluegill in
      control aquaria and  in aquaria treated with
      zinc in a 7-day experiment	75
                                   ix

-------
                                    TABLES

Number                                                                   Page

   1  Fish Species at the Columbia  Generating  Station Site  	  11

   2  Water Temperature (°C) at Various  Sites  in  the  Spawning Marsh . .  21

   3  Size (in Millimeters) of  Individual  Northern Pike Fry
        Captured in Rocky Run Slough,  Spring  1976 	  25

   4  Water Quality Data for Various  Stations  in  the  Ilarsh—
        March 21-May 12, 1977	28

   5  Concentrations of Trace Elements  in  Northern Pike Eggs and Fry
        Used in the In Situ Bioassay,  Spring  1977	31

   6  Water Chemistry for 1978  Laboratory  Experiment	33

   7  Relationship of Construction  Activities  to  Pike Year-Classes. . .  45

   8  Vegetation Types for Various  Wetlands  Located Along the
        Wisconsin River 	 .........  46

   9  Chemical Characteristics  of Madison, Wis.,  Tap
        Water in which Flagfish Were  Raised, and  Dilution Water
        in which Zinc Exposures Were  Conducted	57

  10  Nominal Zinc Concentration  (ppm),  Length of Exposure (Days),
        and Recovery Time Before  Breeding  (Weeks), for the Zinc
        Exposures of the Parental and  Three  Generations
        of Flagfish	57

  11  Zinc Concentrations for the Parental,  First, Second,
        and Third Generation Exposures	58

  12  Standard Lengths of Survivors and  Nonsurvivors  for
        Parental-Generation Flagfish  Exposed  to 0.8 mg/Liter Zinc
        for 17 Days	60

  13/  Standard Lengths and Wet  Weights  of  Parental-Generation
        Flagfish Used to Produce  the  First Generation 	  60

  14  Spawning Data for Parental-Generation  Flagfish  	  61

-------
 15  Lengths and Wet Weights of First-Generation  Fish  Used
       for Spawning  	 .........   62

 16  Summary of Spawning Data for F^-F^ Generations  of Flagfish  ...   63

 17  Number of Parents Contributing Larvae for Each  Generation  ....   68

 18  Routinely Determined Characteristics of Water Used in  the
       Temperature-Preference Bioassay ... 	   72

 19  Zinc Concentrations (ppm) of Water in Treatment and Control
       Aquaria 	  .....   73

 20  Zinc Tissue Concentrations (ppm) at the End  of  the
       Experiment for Randomly Selected Fish from Treatment
       and Control Tanks ....... 	   76

A-l  Number of Fish Caught at Each Sampling Station, 1976	85

A-2  Number of Fish Caught at Each Sampling Station, 1977	86

A-3  Number of Fish Caught at Each Sampling Station, 1978	87

B-l  Marshes Near the Columbia Generating Station 	   88

D-l  Summary of pH Effects on Aquatic Organisms	97
                                   XI

-------
                               ACKNOWLEDGMENTS

     We are grateful to numerous students  and workers  at the Laboratory of
Limnology of the University of Wisconsin-lladison  for assistance in the field
and the laboratory.  In particular, we  recognize  Jeffery Boxrucker, Douglas
Stamra, Steven Voss, Walter Gauthier,  Samuel  Sharr,  Jane Hillstrora, and
Kathryn Webster.  Warren Buchanan  interpreted the infrared aerial wetland
photographs.  Dr. Philip Helmke and his staff at  the University of
Wisconsin-lladison Soil Science Department  analyzed  the trace-metal content
of northern pike fry.  Steve  Horn  prepared many of  the figures, including
the wetland vegetation maps.  Stephanie Brouwer and Kari Sherman deserve
special mention, for their efforts  in  editing this report.

     Researchers responsible  for the  three major  sections of this report,
under the direction of Dr. John J. Magnuson, were Frank J. Rahel, Michael J.
Talbot, and Anne 11. Forbes (Effects of  the Columbia Generating Station on
Fish Spawning); Frank J. Rahel (Selection  for Zinc  Tolerance Over Four
Generations of Flagfish); and Patricia  A.  Medvick (Use of Temperature
Preference and Activity as a  Sublethal  Bioassay for the Toxic Effects of
Zinc to the Bluegill).  The literature  reviewed in  the appendices were
prepared by Anne 11. Forbes, Walter A.  Gauthier, Dorothy 11. Harrell, and
Frank J. Rahel.
                                      xii

-------
                                  SECTION 1

                                 INTRODUCTION
     Coal-fired electric generating stations play an important role  in
energy production in the United States and are likely to increase in
importance as other fossil fuels become scarce (Gordon  1978).  Increased
coal use has serious environmental consequences, and the changes resulting
from power-plant construction and operation are now receiving considerable
attention nationally (Glass 1978).  Impact studies have been conducted  to
predict and, we hope, to mitigate the negative impacts  of coal-fired
generating stations on both terrestrial and aquatic ecosystems.  One of
these stations, the Columbia Generating Station near Portage, Wis.,  has been
the focus of an extensive 3-yr study funded by the U.S. Environmental
Protection Agency (EPA).  As part of this multidisciplinary effort,  our
research has sought to assess the effects of the station on the local fish
populations.  The study has three components:  (1) a field study to
determine the importance of the generating station site as a spawning ground
for fish and to assess the station's effects on fish reproduction;  (2)  a
laboratory study to determine if fish populations can rapidly evolve
tolerance to trace-element contaminants released by coal-fired generating
stations; and (3) a laboratory study to determine whether a bioassay
utilizing temperature preference and activity could detect changes  in fish
behavior after exposure to sublethal levels of trace elements.

     The Columbia Generating Station is located on wetlands near the
Wisconsin River in south-central Wisconsin (Figure 1).  The 1,104-ha site
contains a wide range of plant and animal communities and includes  aquatic,
wetland, and forested areas. The site is bordered by Duck Creek on  the
north, Rocky Run Creek on the south, and the Wisconsin  River on the  west.
Of the original acreage, one-half has been altered permanently by the
installation of a 203-ha cooling lake, 28-ha ash basin, coal-handling
facilities, and various other structures.  The station has two power-
generating units, Columbia I and Columbia II.  Construction of Columbia I, a
527-MW unit with a 152-m boiler chimney equipped with two hot-side
electrostatic precipitators, began in 1971; operation began in April 1975.
Cooling water for the unit is recycled continuously through the cooling
lake. Columbia II, a unit of similar size but with a 198-m stack and sulfur-
removal scrubbers, began operation in March 1978.  Cooling towers were  built
to remove excess heat from Columbia II to minimize further thermal  loading
to the cooling lake.  Fly ash and bottom ash produced during operation  of
Columbia I are pumped as a slurry into the ashpit where the ash particles
settle out in a series of lagoons.  The water is then pumped to the  ashpit
drain and eventually reaches Rocky Run Creek.  Columbia II adds bottom  ash
to the ashpit; all fly ash from Columbia II is disposed of dry.

                                    -1-

-------
                  •Vi
Figure 1.  Map of the study site at Columbia  Generating Station.
                                     -2-

-------
      A  diverse  fish  community lives in the remaining wetlands, in creeks
bordering  the site on  the  north and south, and in the Wisconsin River west
of  the  site.  Wetlands on  the station site are spawning and nursery grounds
for  important game fish, including  northern pike (Esox lucius), muskellunge
(Esox masquinongy),  and walleye (Stizostedion vitreum vitreum) (Ives and
Besadny 1973).   Four factors  related to construction and operation of the
generating station could potentially affect resident fish populations:
habitat  loss, habitat  alteration, cooling lake intake, and acid
precipitation.

      About 446  ha  of the original  1,104-ha site have been altered by
construction of  the  facility.  Much of the habitat  lost, including land used
for  the cooling lake,  was  formerly  sedge meadow, hence ideal spawning ground
for  game fish (Priegel and Krohn 1975).  Since fish that use these spawning
grounds  are part of  the Wisconsin River population, negative effects on the
station  site could affect  the Wisconsin River fishery.  The magnitude of
these effects would  depend on the availability of alternate spawning sites
and  the  possible involvement  of unique homing stocks.

      Habitat alterations in the remaining wetlands  continue because of
increased  ground-water discharge and waste heat from the cooling lake
(Bedford 1977).  Before construction of the facility, the upland sloped
gradually  to the flood plain  sedge  meadow and the Wisconsin River.  Ground-
water flow of 1  ft /s  from adjacent uplands maintained the sedge meadow
water levels. The flow varied seasonally,  being high during spring and lower
during  the summer.   The cooling lake established a  9-ft hydrostatic head
above water levels in  the  adjacent  wetlands and drastically altered the
natural  ground-water pattern  (Figure 2).   Now warm  water from the cooling
lake  at a  flow  of 4  ft /s  seeps west into the sedge neadow.  The seepage of
warm ground water into the rooting  zone of plants and the associated rise in
water level resulted in dramatic vegetation changes.  The area west of the
cooling  lake, formerly dominated by perennial sedges, is being transformed
into  a  community dominated by annuals  and hydrophytic perennials such as
cattails.   Sedges offer dense mats  of  vegetation for spring spawning fish,
but  the annuals  and hydrophytic perennials generally die down each winter
and  provide little suitable spawning substrate.

      The warm-water seepage into the meadow west of the cooling lake is
masked  by  volumes of Wisconsin River flood water in spring and does not
directly affect  spawning fish or eggs.   However,  the ashpit drain (east and
south of the cooing lake)  also receives cooling lake seepage and remains
several degrees  warmer  even in the  spring.

      Habitat alterations in the wetlands  south of the site along Rocky Run
Creek are  also caused  by effluent from  the ashpit.   Metal oxides that
compose the major reactive portions  of  the ash cause the pH of the ashpit
water to rise 10 to  11  (Andren et al.  1977).   Since Wisconsin water quality
standards  prohibit the  release of water at a pH above 8, sulfuric acid is
added before the ash effluent  is discharged.   The addition of acid causes
the  precipitation of elements  such  as  barium and aluminum into a floe that
coats the  bottom of the ashpit  drain and  flows  into Rocky Run Creek.   Thus,


                                     -3-

-------
            River
      UJ
            River    *-Studv Area
                                      Cooling  Lake
      CO
      UI
                         500
     1000
METERS
1500
2000
Figure 2.  Ground-water flows before and after construction of the
           Columbia cooling lake.  Arrows represent integrated flows
           1 m3/min, normal to the east-west cross section along  the
           length of the cooling lake (Andrews and Anderson 1980).
                                    -4-

-------
dissolved and particulate trace elements, fly-ash particles, and  perhaps
organic contaminants in the ash effluent continually enter these  streams,
which then  flow into the extensive Rocky Run wetland area.  Many fish
species, including the sensitive early life-history stages of important game
fish, live in these waters and are exposed  to the effluent.

     Although the generating station recycles cooling water  through  the
cooling lake, evaporative losses are made up by using Wisconsin River
water.  Since water and any organisms withdrawn from the  river are not
returned, the plant is analogous to a predator.  Fish loss due to
impingement and entrainment depends on the  volume of water removed,  the
patchiness of fish distribution, and the ability of fish  to  avoid
entrainment.  A 1-yr study of egg and larval fish entrainment and juvenile
and adult fish impingement at the Columbia  site (Swanson  Environmental,  Inc.
1977) reported insignificant numbers of fish losses.  The total river flow
removed by the intake water at Columbia presently averages 0.3%,  with a
maximum of 1.08%.  As long as the Columbia  intake continues  to remove a
small percentage of the river flow, we expect no measurable  effects  of
entrainment on the river system.  An exception might occur when organism
distribution is patchy near the intake, and a significant portion of one
year-class (e.g., walleye larvae) is entrained.  Aside  from  acting as a
predator by removing organisms from the Wisconsin River,  the usual types of
entrainment effects (mechanical, toxic, and thermal) do not  apply to the
Columbia station, since organisms and water are not returned to the  river.

     Loss of fish populations due to lake acidification is related directly
to acid precipitation from fossil-fuel combustion (Gorham 1976).   The
likelihood of waters undergoing acidification depends on  the edaphic
characteristics of their drainage basins and the intensity of acid input.
Waters near the Columbia station site have  a high buffering  capacity because
the drainage basin is calcareous.  Although the problem has  not been studied
directly at the Columbia site, predictions  based on the extensive literature
indicate little potential for acid precipitation damage to the aquatic
systems near the Columbia station.

DESIGN OF THE FISH STUDY

     The impacts of the four above factors  on fish populations were
investigated through both the field and laboratory components of  this study,
as well as by reference to existing literature.  Some of  the concerns are
site-specific; others are more general and  therefore applicable to other
locations. Overall conclusions are presented in Section 2.

     Section 3 documents the site-specific  effort to determine the
importance of the generating station as a spawning ground for northern pike,
muskellunge,  and walleye.  Spring sampling  of these populations yielded
information on the use of the marsh for spawning as well  as  on year-class
strengths related to both natural and artifical changes.  The effect of the
ashpit effluent on fish reproduction is discussed in Section 3, as is the
relative importance of the spawning marsh to the total Wisconsin  River
fishery.
                                     -5-

-------
     Sections 4 and 5 concern questions that may arise at other generating
sites.  Section 4 describes the methods and results of the metal-tolerance
study, in which flagfish were bred in the laboratory for resistance to
zinc.  Section 5 deals with the use of temperature preference and activity
as a bioassay to detect subtle changes in fish behavior after exposure  to
sublethal metal levels.

     Appendices include reviews of the literature on entrainment, acid  rain,
and fly-ash disposal.
                                    -6-

-------
                                  SECTION 2

                                 CONCLUSIONS
     The major factors affecting fish at the  Columbia  site are  habitat  loss
and habitat modification.  Construction of the  Columbia  Generating  Station
eliminated approximately  18% of the shallow-water  sedge  meadow  located
between the Prairie du Sac and Wisconsin Dells  dams.   This habitat, when
inundated by spring floods, is utilized by spawning northern  pike and
muskellunge.  The station site currently contains  about  22% of  the  deep-
water sedge meadow and 0.8% of the shallow-water sedge meadow likely to be
used by spawning northern pike in this section  of  the Wisconsin River.
Tagging efforts indicated that northern pike  from  as far away as Lake
Wisconsin (17 km downstream) migrate to the Rocky  Run wetlands  to spawn.
However, these wetlands continue to be affected by effluent from the ashpit
drain and by underground seepage of warm water  from the  cooling lake.

     Northern pike, muskellunge, and walleye  spawned in  areas affected  by
ashpit effluent; in fact, northern pike were  apparently  attracted to the
ashpit drain because of warmer spring water temperatures and  higher current
speeds.

     The presence of ripe and spent northern  pike  adult  spawners in 1976-78
indicated that spawning occurred in areas affected by the ashpit effluent.
The presence of newly hatched fry in 1976 indicates that reproduction was
successful.

     Analysis of northern pike year-class strengths suggested that  the  1976
year-class (the first year-class affected by  the plant's operation) may be
reduced.  Further monitoring of population structure is  warranted since
year-classes hatched after the station began  operation are now  reaching
maturity and should be returning to spawn.

     In a laboratory bioassay, hatching success of pike  eggs  incubated  in
unfiltered ashpit drain water was lower than  for eggs raised  in filtered
ashpit drain water; therefore, the flocculent precipitate found in  ashpit
drain water appears to hinder pike egg development.  When the flocculent
precipitate was removed by filtering, eggs hatched equally well in  ashpit
drain water and Rocky Run Creek water.

     Fry hatched in the ashpit drain contained  elevated  levels  of only  one
element, sodium, compared to fry hatched at other  locations in  the  marsh.
Acute toxicity due to trace-element bioaccumulation is unlikely to  be a
problem for fish eggs or newly hatched fry in the  ashpit drain.
                                     -7-

-------
     In the metal-tolerance study the selected population had a higher
resistance to zinc after the first two generations, but did not differ from
the unselected population in the third generation.  Possible explanations
for the failure of selection to continually increase zinc tolerance include
inbreeding depression and carry-over effects passed from mother to offspring
through the egg cytoplasm.  After 2 to 8 weeks recovery time, flagfish which
showed a zinc exposure lethal to the majority of the population, reproduced
as successfully as unexposed fish.

     The temperature preference and activity apparatus tested in the
sublethal zinc bioassay is no more sensitive than bioassay methods used by
previous investigators.

     Potential damage to Wisconsin River fish and invertebrate populations
from entrainment or impingement on water intake structures appears minimal.

     Acid precipitation is not considered a potential problem for aquatic
ecosystems at the Columbia site because of the high hydrogen-ion buffering
capacity resulting from the calcareous nature of the drainage basin.
                                     -8-

-------
                                   SECTION  3

         EFFECTS OF THE COLUMBIA GENERATING  STATION  ON  FISH SPAWNING
INTRODUCTION

     Construction of a coal-fired  electric  generating  station on a flood
plain of the Wisconsin River  (Columbia  County, Wis.) has  resulted in
alteration of an important  fish-spawning  habitat.  The  station site formerly
contained extensive wetland areas,  particularly  during spring floods when
many fish species migrate to  such  areas to  spawn on inundated vegetation.
Among those species known to  have  used  the  Columbia Generating Station site
for spawning are northern pike  (Esox  lucius), muskellunge (Esox
masquinongy), and walleye (Stizostedion vitreum  vitreum). Construction of
the station permanently altered about one-half of  the  original 1,104 ha;
much of the wetland affected  was sedge  meadow, an  ideal spawning habitat for
these fish.  Portions of the  remaining  wetlands  and Rocky Run Creek have
undergone physical and chemical changes that may also  influence fish
reproduction.  Given these  considerations we undertook a  site-specific study
to determine both the importance of the station  site as a spawning ground
for Wisconsin River game fish and  the effect of  the power plant on the
reproductive success of these fish.   Our  study involved the  following:  (1)
netting on the site to discover which areas were important spawning grounds;
(2) both -in situ and laboratory bioassays to assess the effects of ashpit
effluent on hatching and larval survival; (3) aging of scale samples to
relate year-class strengths to both natural and  artificial changes in the
environment; and (4) use of infrared  aerial photography to compare the
spawning habitat at the station site  to the total  spawning habitat available
in this section of the Wisconsin River.

     One of the most obvious  effects  of the generating station is the
introduction of ashpit effluent into  potential spawning areas.  This
effluent is pumped from the ashpit  settling basins  into the  ashpit drain.
The ashpit drain joins a creek that flows through  portions of the wetlands
adjoining the site and then enters  Rocky  Run Creek  1.5 km above its mouth at
the Wisconsin River (Figure 1).  The  effluent waters contain elevated levels
of some trace elements and other coal-combustion by products (Andren et al.
1980, Helmke et al., Unpublished).  Beginning in January  1977 sodium
bicarbonate was routinely added to  the  pulverized  coal to increase the
efficiency of the electrostatic precipitators.   This treatment resulted in
increased conductivity in the ashpit  drain and Rocky Run  Creek and, in fact,
served as a useful tool for measuring ash-effluent  concentration downstream
from the generating station.  The water upstream of the ash  effluent in the
mint drain and in Rocky Run Creek  is  usually higher in alkalinity, hardness,
and pH and lower in turbidity (Magnuson et al. 1980).

                                    -9-

-------
     Invertebrate populations in the ashpit  drain  decreased  in abundance and
species composition after generating station operation  began (Magnuson et
al. 1980).  Schoenfield (1978) demonstrated  a  tenfold  increase in chromium
and barium concentrations in several taxa of these  ashpit-drain
invertebrates (especially Asellus and  Hyalella)  since  the  start of station
operation.  The ashpit drain itself has a very sparse  resident fish
population (ictalurids and some cyprinids) probably as  a  result of
inhospitable chemical and physical conditions as well  as a lack of food
organisms.  Studies of another ashpit-drain  system also documented a
depauperate invertebrate fauna and the presence  of  only one  fish species
(Gambusia affinis)  (Cherry et al.  1976).

     The Rocky Run  Creek area has a diverse  fish community and is an
important site for  northern pike, muskellunge, and walleye.   These species
spawn on flooded wetlands where the newly hatched  larvae  remain for up to
several months before emigrating to nearby rivers  or lakes.   A
preoperational impact statement completed by the Wisconsin Department of
Natural Resources (Ives and Besadny  1973) documented the  occurrence of 47
fish species on or  near the station site  (Table  1).  Included was one
species listed as threatend in Wisconsin, the  mud  darter  (Etheostoma
aspirgene).  Eggs and fry of both walleye and  northern pike  were also
collected on the station site during the  preoperational study.

METHODS

     We conducted extensive fyke netting  on  flooded wetlands accessible from
Rocky Run Creek from  1976 to 1978 to determine which areas were important
spawning grounds, to  see if fish bypassed the  station site  to spawn, and to
assess year-class strengths of northern pike,  the  predominant species.  By
tagging fish we hoped to document  the  areas  from which fish  traveled to
reach the spawning  grounds and to determine  whether homing to a certain site
was recurring in successive years.

     After discovering that pike were  spawning in  areas receiving ashpit
effluent, we compared reproductive success in  those areas  with reproductive
success in nearby unaffected control areas.  Collection of pike fry in the
field to study the  natural hatch was unsuccessful; hence,  we conducted both
an in situ and a laboratory bioassay to assess the effects of ashpit
effluent on egg hatching and larval  survival.  Water from both affected and
unaffected areas was  used, and survival was  monitored daily.  Northern pike
fry hatched in various locations in  the marsh  were analyzed  for trace
elements to assess  uptake in areas affected  by the ashpit effluent.

     After spawning,  fish return to  a  63-km  section of the Wisconsin River
bordered upstream by  a dam at Wisconsin  Dells  and  downstream by a dam at
Prairie du Sac which  forms Lake Wisconsin.  During  spring  floods in this
stretch of the river, many other backwater  areas are formed  that could serve
as pike spawning grounds.  The  relative  importance of spawning habitat on
the plant site to the total  spawning habitat available in this section of
the river was determined by  infrared aerial  photography.   Photo-
interpretation allowed us to classify  wetlands into various  types according
to their value as pike-spawning grounds and  to calculate  the acreage of each


                                     -10-

-------
       TABLE  1.   FISH SPECIES AT THE COLUMBIA GENERATING STATION SITE
  Common name
      Scientific name
Collection
   sites3
Reference
Bigmouth buffalo
Black bullhead
Black crappie
Blacknose shiner
Blackstripe topminnow
Bluegill
Bluntnose minnow
Bowfin
Brook silversides
Brown bullhead
Brown trout
Bullhead riinnow
Carp
Central mudminnow
Channel catfish
Chestnut lamprey
Emerald shiner
Fathead minnow
Flathead catfish
Freshwater drum
Golden redhorse
Golden shiner
Grass pickerel
Green sunfish
Johnny darter
Lake sturgeon
Largemouth bass
Mooneye
Mud darter
Muskellunge
Ictiobus cyprinellus
Ictalurus melas
Pomoxis nigromaculatus
Notropis heterolepis
Fundulus notatus
Lepomis macrochirus
Pimephales notatus
Amia calva
Labidesthes sicculus
Ictalurus nebulosus
Salmos trutta
Pimephales vigilax
Cyprinus carpio
Umbra limi
Ictalurus punctatus
Ichthyorayzon castaneus
Notropis atherinoides
Pimephales promelas
Pylocictis olivaris
Aplodinotus grunniens
Moxostoma erythrurum
Notemigonus chrysoleucas
Exox americanus vermiculatus
Leponis cyanellus
Etheostoma nigrum
Acipenser fulvescens
Micropterus salmoides
Hiodon tergisus
Etheostoma aspirgene
Esox masquinongy
      C             2
    B,C,D          1,2
  A,B,C,D         1,2
      B             1
      B             1
  A,B,C,D         1,2
      B             1
    A,B,D          1,2
      B             1
     B,D           1
      C             2
      B             1
  A,B,C,D         1,2
      B             1
     A,B           1
      C             2
      B             1
      B             1
      A            1,2
  A,B,C,D         1,2
     B,D           1
     B,D           1
      B             1
      B             1
      B             1
      A             1
      B            1,2
      A             1
      B             1
    A,B,C          1,2

     (continued)
                                     -11-

-------
TABLE 1 (continued)

Common name
Northern pike
Pirate perch
Pumpkinseed
Quillback
Rainbow trout
Redhorse
Red shiner
Rock bass
Sand shiner
Sauger
Shovelnose sturgeon
Smallmouth bass
Smallmouth buffalo
Spotfin shiner
Spotted sucker
Tadpole madtom
Walleye
White bass
White crappie
White sucker
Yellow bullhead
Yellow perch



i
Scientific name
Esox lucius
Aphredoderus sayanus
Lepomis gibbosus
Carpiodes cyprinus
Salmo gairdneri
Moxostoma sp.
Notropis lutrensis
Ambloplites rupestris
Notropis stramineus
Stizostedion canadense
Scaphirhynchus platorynchus
Micropterus doloraieui
Ictiobus bubalus
Notropis spilopterus
llinytrema melanops
Noturus gyrinus
Stizostedion vitreum vitreum
Morone chrysops
Poraoxis annularis
Catostomus coranersoni
Ictalurus natalis
Perca flavescens



Collection
sites3
A,B,D
B,C
A,B,D
B
C
C
B
A,B,D
B
A,B,C,D
A
A,B
C
B
B
B
A,B,C,D
A,B
A,B,D
B,D
A,B,D
A.B.C.D



Reference
1,2
1,2
1,2
1
2
2
1
1,2
1
1,2
1
1
1
1
1,2
1
1,2
1,2
1,2
1,2
1,2
1,2


aCollection sites include:  A—Wisconsin  River,  B—Duck  Creek,  C—Rocky Run
 Creek, and D—On-site flooded areas.
 List compiled by investigators from  (1)  the Wisconsin Department  of
 Natural Resources and (2) the University of Wisconsin-Iladison.
                                    -12-

-------
type.  By comparing various watersheds we could determine  what  percentage of
the total available pike spawning habitat in that section  of  the  Wisconsin
River was being influenced by the generating station.

     Scale samples from pike, muskellunge, and walleye were aged  in  an
effort to relate year-class strengths to both natural and  man-caused changes
in the environment.  The introduction into spawning areas  of  the  ashpit
effluent, which had a complex and largely unknown chemical make-up,  was  one
such change.  Construction activities, which began in January 1971,
destroyed a large portion of the sedge meadow and undoubtedly influenced
spring flood patterns.  Studies have shown that flooded vegetation,
preferably dense mats of sedges (Carex sp.), is necessary  for successful
pike and muskellunge reproduction (McCarraher and Thomas  1972,  Priegel and
Krohn 1975).  Since the generating station began operation, seepage  under
the cooling lake dikes has modified water temperatures and vegetation types
in the adjacent wetlands (Bedford 1977).  The general trend has been toward
replacement of shallow-water marsh dominated by the sedge  (Carex  lacustris)
and other perennials to deep water dominated by annuals and hydrophytic
perennials such as arrowhead (Sagittaria sp.) and cattail  (Typha  sp.).
These changes reduced the amount of densely matted vegetation available  for
spring spawning fish.  Finally, one natural factor that strongly  influences
reproductive success is the timing and extent of spring flood levels.
Johnson (1956) found a direct correlation between high spring water  levels
followed by a small decline in levels during egg incubation and production
of a strong northern pike year-class.

     With the above considerations in mind, our overall goal  was  to
determine the importance of the generating station site as a  spawning ground
for spring-spawning game fish and to assess the plant's effects on the
reproductive success of these fish.  More specific information  on methods is
included in the following sections.

Survey of Spawning Grounds

     To sample fish species that spawn on the site, fyke-net  sites were
established at five strategic locations (Figure 3) and checked  regularly
from ice-out in late February through the end of the spawning season in  late
April.  All fish captured during upstream movements were released on the
upstream side of each net.  A description of each site and the  years that
the nets were worked follows.

     Net 1, worked between 1976 and 1978, was located on Rocky  Run Creek at
the bridge on County Trunk Highway JV.  This site, upstream from  the plant,
was unaffected by construction or operation of the generating station.  This
net completely crossed the river channel, except for peak  floods,  and caught
fish that passed by the station site.  Water depths ranged from 1.5  to 3 m,
depending on flood stage, and the current was generally strong.   Spawning
habitat was available immediately above the net when the creek  flooded
surrounding marshland.

     Net 2, worked between 1976 and 1978, was located in Rocky  Run Creek
downstream of its confluence with the ashpit drain.  The net  was  placed  in a

                                   -13-

-------
         1.5    km
Figure 3.  Location of fyke nets in spawning marshes on the Columbia site.
           (* on insert shows the location of the Columbia Generating
           Station in south-central Wisconsin.)

                                   -14-

-------
 channel  between two  old  bridge abutments and,  except for peak flood periods,
 captured most  fish entering  the marsh system.   Water depth ranged from 1.5
 to  3  m depending on  flood stage, and the current was moderate.  Fish passing
 this  point might spawn in the  habitat immediately above the net, or they
 might proceed  to spawning areas near nets 1,  3, or 4. Ashpit effluent was
 well  mixed with Rocky Run water by the time it reached net 2.

      Net 3,  worked between 1976 and 1978, was  located at the northern end of
 the Rocky Run  Creek  backwaters and in the main current channel leading into
 the sedge meadow near the cooling lake.  This  net caught fish that migrated
 into  wetlands  beyond those affected by ashpit  effluent.  Passage around the
 net was  often  possible when  the entire area was flooded, and generally water
 levels dropped  sharply at this site after the  spring floods.  Although
 spawning habitat was available near the net,  it was mainly located upstream
 in flooded sedge meadows.

      Net 4,  worked between 1976 and 1978, was  located in the ashpit drain
 above  its junction with  Rocky  Run Creek in a diked channel that caught most
 of the fish  migrating upstream.   This net caught fish entering spawning
 areas most affected  by ashpit  effluent.   The current was moderate and water
 levels were  generally 0.5 to 2 m.   Unlike net  site 3, this site retained a
 flow  of  water  throughout  the year.   Abundant spawning habitat was available
 below and above this net. Clumps of the precipitated floe described earlier
 were  often found floating downstream,  and finer floe particles could be
 observed in  the water column.

      Net 5,  worked in 1978,  was  located above  the confluence of the mint and
 ashpit drains,  and was used  to determine if  fish would migrate all the way
 up the ashpit  drain  to reach spawning ground unaffected by ashpit
 effluent.  It  completely  sealed  off a small  channel that was usually less
 than  1 m deep  but contained  water  throughout the year.

      The Rocky  Run slough areas  were electroshocked in 1976 and 1977 to
 determine if many fish were  avoiding the nets,  but still utilizing the site
 for spawning.   Areas downstream  from net 2 were shocked also to determine
 their importance as spawning grounds.   Because  of low water levels and
 technical problems, we found that  our  fyke nets were a more effective method
 of collecting  fish.

     Northern pike, muskellunge,  and walleyes  were measured, sexed,  and
 fitted with a monel tag,  from  the  National Band and Tag Co.,  inserted into
 their preopercular bone.   The  tag  carried an identification number as well
as the label "University  of  Wisconsin-Madison."  To determine  the
distribution of  these fish during  the  nonspawning season,  we posted  notices
at various  boat  landings  and fishing areas  offering a reward to anglers who
returned tags from fish they caught.

     To  estimate spawning  success  on the plant  site,  we set fry traps in
 1976 and 1977 in areas corresponding  to  net sites  2,  3,  and 4.   The  traps
were generally  ineffective,  as were  light traps tried in 1978.   After study
of the natural hatch proved  unfeasible,  we conducted  field  and  laboratory


                                    -15-

-------
bioassays to assess the effect of the ashpit effluent on northern  pike  egg
and fry survival.

Effects of Ashpit Effluent on Reproductive Success

     After documenting the use of areas receiving ashpit effluent  as
spawning grounds in 1976, we conducted an in s~itu bioassay  in  1977 to
determine if the effluent was affecting survival of eggs and larvae of
northern pike, the predominant game species in the area.  Fertilized eggs
were obtained from the Wild Rose Fish Hatchery of the Wisconsin  Department
of Natural Resources.  Eyed eggs were incubated in the main stream channel
at three locations:  (1) in the ashpit drain; (2) above the plant  site  in
Rocky Run Creek; and (3) downstream Rocky Run Creek after ashpit effluent
water was mixed well with creek water.  These locations corresponded to net
sites 4, 1, and 2, respectively.

     Ten incubation bottles, each containing approximately  250 eggs, were
anchored at each location.  Incubation bottles were 1-gal plastic  containers
with plastic screen on the sides and bottom, current deflectors  to maintain
water exchange, and floats to keep them at the surface.  Each  day  we emptied
accumulated silt, counted live eggs and larvae, and observed the stage  of
egg development.  Dead eggs were removed to minimize disease.  Water
temperatures at each station were recorded continuously with a Ryan
Instruments temperature recorder, and relative siltation loads were
estimated from silt deposition in graduated cylinders over  24-h  periods.
Dissolved oxygen, conductivity, water current speeds, pH, alkalinity, and
EDTA hardness were measured regularly with standard techniques (American
Riblic Health Association et al. 1975).  Survival was calculated daily  as
the percentage of stocked eggs still as eggs or fry.  The experiment was
repeated with newly hatched pike larvae, also obtained from the  Wisconsin
Department of Natural Resources.

     Northern pike fry that hatched in our incubation bottles  were analyzed
for 20 trace-element concentrations by neutron activation.  The  analyses
were performed by the Trace Elements Subproject under the direction of  Dr.
Riillip Helmke (Helmke, unpublished).  Fry hatched in areas affected by the
ashpit effluent were compared with fry from unaffected waters.  Since the
ashpit effluent was known to contain elevated levels of certain  trace
elements, we felt it important to document any biological accumulation  by
this sensitive life-history stage.

     At all three locations, the incubation bottles accumulated  silt, but
the problem was greatest at the Rocky Run stations where the daily
accumulation of fine dark silt often completely covered the eggs.   Although
sediment also accumulated in the ashpit drain bottles, most of it  consisted
of the semibuoyant chemical floe.  Ashpit drain eggs remained  amber in
color, whereas Rocky Run eggs were coated with brown silt.  In nature,  eggs
are spawned over a mat of sedge grasses and do not become buried by
sediments.  Therefore, we repeated the experiments in  1978  under controlled
laboratory conditions where siltation effects could be eliminated.
                                    -16-

-------
     Egg cups made from polyinylchloride (PVC) piping covered at  the  bottom
with plastic screen were used to hatch eggs in a laboratory chamber
maintained at 12°C.  Ten cups of approximately 100 eggs each were  incubated
in individual 500-ml beakers containing ashpit drain water, Rocky  Run water,
or the naturally occurring mixture.  A similar set of egg cups was exposed
to water from these same sources that had been passed through a 3-ym  filter
to eliminate sediment or floe, or both.  Each lot of eggs was given a fresh
exchange of their respective test water daily, and dead eggs and  fry  were
removed.  The experiment proceeded until all eggs either hatched  or died.

Year-Class Strengths

     Northern pike, muskellunge, and walleye were aged by counting annuli on
their scales (Williams 1955).  Scales were collected from above the lateral
line below the dorsal fin. After cleansing in a solution of 0.1 N  NaOH,
impressions of the scales were made on cellulose-acetate slides.   These
impressions were then projected onto a screen, and the annuli were
counted.  Year-class strengths were then related to factors such as water
levels or site construction activities that might have influenced
reproductive success.

Importance of the Station Site to the Wisconsin River Fishery

     An inventory of potential pike spawning areas in the section  of  the
Wisconsin River bordered by the Wisconsin Dells and Prairie du Sac dams was
prepared by manual photo-interpretation of 70-mm color infrared
transparencies taken from an altitude of 11,000 ft.  A series of  180  images
along 12 flight lines was taken during low water levels on 9 September  1977
at wetland at a .scale of 1:62,500.  Flights took place in the fall when
maximum vegetation was exposed.

     The first step in the analytical process was preparation of  a base map
on which to trace the photo-interpretations.  Because the photoscale
approximated that of the standard U.S. Geological Survey (USGS) (15-min)
topographic quadrangle maps, transparent mylar copies of the USGS  maps were
used as base maps.  Each photograph was placed under the transparent  base
map, and vegetation areas were transferred to the base map.

     An experienced photo-interpreter identified the resources on  the
imagery.  Ground checking of mapped wetland classes showed an excellent
correlation between photo-interpretation and actual vegetation types. From
the infrared photographs the following wetland classifications were
delineated:

     1)  Deep-water sedge meadow.  The deep-water sedge meadows consist of
lake sedge (Carex lacustris), bullrush (Scirpus spp.), and bluejoint  grass
(Calamagrostis canadensis and C^. inexpansa); some colonies of cattail (Typha
sp.) are mixed in.  Other floating and emergent aquatic macrophytes that
often occur in deep-water sedge meadows include arrowhead (Sagittaria spp.)
and burreed (Sparganium eurycarpum).  A deep-water sedge meadow is
characterized by inundation throughout the year.
                                    -17-

-------
     2)  Shallow-water sedge meadow.  This category is dominated  largely  by
either tussock sedge (Carex stricta) or bluejoint grass.   Forbs such  as
aster  (Eupatorium spp.) are occasionally prominent.   Cattail  is fairly
common, as is reed canary grass (Phalaris arundinacea).   Shallow-water sedge
meadow, as the name implies, has a drier moisture regime  than deep-water
sedge meadow.  For long periods during the growing season  the water table is
somewhat below the surface.

     3)  Emergents and deep-water floating macrophytes.   This category
includes deep-water species floating or barely emerging above the  surface.
Floating mats of duckweed (Lemna spp.) frequently cover stagnant  open water
between other vegetation. Large mats of arrowhead (Sagittaria spp.),  water
lily (Nymphaea spp.) and (Nuphar spp.), or pondweed (Potamogeton  spp.) often
compose this vegetation type, which is often associated with  the  borders  of
open bodies of water.

     4)  Cattail.  Monospecific stands of cattail are separated on the maps
if large enough to outline.

     5)  Reed canary grass.  This densely grown monospecific  vegetation type
often occurs along water courses or in regularly shaped plantations.

     6)  Bluejoint grass.  This species seldom occurs in  large pure stands
and is an indication of a fairly undisturbed wetland  community.

     7)  Drained wetland.  A growth of weedy forbs and shrubs often
indicates ditching and desiccated conditions.  Typically,  remnants of the
original vegetation type are found under the weeds.

     8)  Shrub carr and swamp.  Dogwood (Cornus stolonifera),  spiraea
(Spiraea alba), or willow (Salix Interior) are the most common constituents
of shrub carr.  Lowland tree species of willow, silver maple  (Acer
saccharinum), aspen (Populus tremuloides), or green ash (Fraxinus
pennsylvanica) are commonly interspersed or occur as  solid stands  of  lowland
forest.

     9)  Mixed vegetation.  Vegetation complexes too  detailed to  separate by
individual boundaries were combined and labeled by predominant types. For
example, complexes of shrub-cattail-sedge meadow were fairly  common.

     Areas for each vegetation class were quantified  bu using a Hewlett
Packard Model 9107A digitizer and calculator.  Comparison of  pike spawning
areas  (viz., deep- and shallow-water sedge meadows) on the plant  site with
other  potential spawning areas allowed us to draw some conclusions about  the
importance of the plant site to the Wisconsin River fishery.
                                     -18-

-------
RESULTS

Survey of Spawning Grounds

     Our fyke nets caught a  total  of  21  species  of  fish over a 3-yr period
(Appendix A).  Northern pike was the  dominant  game  fish,  but walleye,
muskellunge, and largeraouth  bass were caught  along  with many carp and
various catostomid species.  Trie average  yearly  catch (Figure 4)  was largest
at the entrance to the Rocky Run slough  area  (net  2)  and smallest in upper
Rocky Run (net 1). No fish were caught in the  mint  drain (net 5).  The low
number of fish caught at nets  1 and  5 indicates  that  most fish entering the
system remain on the station site  to  spawn.   Furthermore, our catches in the
ashpit drain (net 4) were larger than those at the  access point to the sedge
meadow (net 3), indicating that many  northern  pike  move into the  ashpit
drain during spawning migrations.

     Water temperatures at each net  location  indicated that ashpit drain
water was consistently warmer  by several  degrees centigrade than  water
draining the sedge meadow adjacent to the cooling  lake (Table 2).  The
greater number of northern pike spawning  near  the  ashpit drain may be
attributed to the pike's affinity  for warmer  water  currents during migration
(Johnson 1956, Franklin and  Smith  1963).   Fish entering the Rocky Run slough
through main current channels  encounter  an intersection where cooler water
flowing through net 3 mixes with warmer  water  originating in the  ashpit
drain and flowing through net  4.   Preference  for the  warmer water currents
may lead fish close to the ashpit  drain.

     In addition to being warmer,  the ashpit  drain  had a markedly higher
conductivity in 1977 and 1978  than the other  stations. This condition was
due to the use of sodium bicarbonate  in  the coal to increase the  efficiency
of the electrostatic precipitators.   For  example,  February-April  1978
averages for conductivity at 25 C  were 468 ymhos/cm at net 1, 658 at net 2,
805 at net 3, and 1,218 at net 4.   The conductivity decreases downstream,
but is still elevated at the mouth of the Rocky Run slough (net 2).

Effect of Spring Water Levels  on Spawning—

     Spring water levels also  appear  to  influence  the distribution of
spawning fish.  In 1976 and  1978 spring  flood  periods were normal, but in
1977 flooding was greatly reduced  (Figure 5).   Thus,  although large numbers
of spawning northern pike enter the  slough area  each  year, they proceed to
spawning sites near the cooling lake  (net 3)  or in  the ashpit drain (net 4)
only if water levels are high  enough  to  provide  suitably flooded  vegetation.
In 1977 such vegetation was  unavailable,  and  pike  remained in marshy areas
just above the entrance to the Rocky  Run  slough area  (net 2) or moved
further upstream in Rocky Run. This  situation is  indicated by a  reduced
catch per unit effort (= total number of  pike  caught  per number of days net
was set) for nets in the sedge meadow and ashpit in 1977 compared to other
years, but an increased catch  per  unit effort  for  the net in upstream Rocky
Run (net 1) (Figure 6).
                                     -19-

-------
      AVERAGE  FISH  CATCH
                 SEDGE
                MEADOW
          Esocidae (2 spp.)

          Catostomidae (3 spp)

          Other (14 spp.)
                                            ASH PIT
                                              DRAIN
           LOWER
           ROCKY RUN
   UPPER
ROCKY RUN
Figure 4.  Fyke-net catches averaged over the 3-yr period, 1976-78.  Upper
          Rocky Run corresponds to net 1 in Figure 3,  lower Rocky Run to
          net 2, sedge meadow to net 3, and ashpit drain to net  4.  Circles
          are scaled to indicate relative abundance at each net  site.
                               -20-

-------
TABLE 2.  WATER TEMPERATURE (°C) AT VARIOUS SITES IN THE SPAWNING HARSH


1976
March















April








May



1977
March







5
13
15
16
17
18
19
20
21
22
23
24
25
26
27
28
8
10
13
15
20
22
23
28
30
3
5
7
12

2
4
7
9
11
14
Upper
Rocky Run
(Net 1)

2.8

3.4
1.7
2.0
4.7
4.5
4.8
3.4

3.1
6.0
7.0
9.2
6.3

10.4
10.2
8.8
12.5


11.5
9.4
10.1
5.6
12.7
7.7


0.5

4.0


6.0
Lower
Rocky Run
(Net 2)


2.5
3.6
1.3
2.2
5.0
6.6
5.4
3.6
3.6
3.5
7.0
7.2
8.5
6.3
6.8
11.5
11.8
14.2


10.2
12.4
11.0
13.2
6.5
13.5
10.3


0.4
2.0
4.0
6.1
9.4
6.5
Access to
sedge meadow
adjacent to
cooling lake
(Net 3)



1.9
1.0
1.8
3.2
3.6
7.7
3.4
5.0
3.1
5.8
6.1
6.5
4.5
5.1
10.5
9.9
10.6
15.1
11.6
10.4
12.5
10.5
12.1
5.8
13.7
8.4
13.5







Lower
ashpit
drain Hint drain
(Net 4) (Net 5)



8.0
3.7
4.2
7.5
8.6
10.4
3.6
7.5
4.5
7.2
6.7
10.6
6.5
7.7
15.5
11.0
11.4
13.4
12.8
10.1
12.7
10.5
10.4
4.9
14.2
10.2







O • U f — 4. 4 — — . J\
                                 -21-

-------
TABLE 2 (continued)

Upper
Rocky Run
(Net 1)







April


1978
March














April






16
18
21
23
24
28
30
4
5
7

2
3
6
8
10
13
15
17
20
22
24
27
29
30
31
3
6
8
10
12
14
16
6.
4.
5.
6.
5.
8.
8.
3.
4.
10.










5.
1.
8.
8.

13.
7.
6.
7.
8.
8.
11.
11.
5
0
0
0
0
2
0
0
5
0










3
5
0
2

7
9
8
7
8
9
0
0
Lower
Rocky Run
(Net 2)
7
4
6
6
7
9
9
2
3
9

2
1
0
1
0
1
1
1
3
5
2
5
10
6
12
9
7
8
9
8
11
12
•
•
*
•
•
•
•
•
•
•

•
•
*
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
0
0
0
5
7
2
7
2
9
0

8
0
2
0
9
8
0
8
0
0
0
8
1
9
5
9
5
1
7
3
1
0
Access to
sedge meadow Lower
adjacent to ashpit
cooling lake drain Ilint drain
(Net 3) (Net 4) (Net 5)
5.7 10
2.5 7
4.0 7


9.0 10

0.5 3
3.8 6
16.0 14









8
9
3

9.5 11

7.9 16
9.8 11
9.0 8
8
8.9 9
8
12
13
.0
.5
.0


.0

.5
.5
.0









.0
.0
.5

.0

.1
.5
.5
.0
.0
.3
.5
.0

























14.1
8.2
6.5
6.9
8.1



                                    -22-

-------
Figure 5.  Spring water levels in the spawning grounds at the Columbia site during 1976-78.

-------
    7.0

    5.0



    3.0
z
ID
•x.
I
u
u
z
o
o
liO
    0.5
     0.1
    0.05
              1976  77 78

               LOWER

               ROC K Y

                 RUN
                              76  77 78

                            UPPER

                            ROC K Y

                              RUN
  76 77 78

SEDGE

ME A DOW
  76 77  78

A S H PIT

DRAIN
                              LOCATION  AND   YEAR
Figure 6.  Catch of northern pike per unit effort on the Columbia  site during
           1976-78.  Upper Rocky Run corresponds to net 1 on Figure  3, lower
           Rocky Run to  net 2,  sedge meadow to net 3, and ashpit drain to net
           4.  Catch per unit effort was calculated by dividing  the  total
           number  of pike caught by the total number of days a given net was
           worked.
                                     -24-

-------
Capture of Pike Fry—

     A few northern pike fry were captured  in  the  Rocky  Run slough area in
1976.  This indicates that eggs were hatching  and  larvae were  surviving in
these wetlands (Table 3).  About one-third  of  the  fry  were  caught in the
ashpit drain where waters are most affected by ashpit  effluent.   Small
sample sizes precluded any comparisons  of growth rates or estimates of fry
abundance between sampled areas.

       TABLE 3.  SIZE (IN MILLIMETERS)  OF INDIVIDUAL NORTHERN  PIKE FRY
                  CAPTURED IN ROCKY RUN SLOUGH, SPRING 1976


Date
captured
5 May
7 May
10 May
12 May
14 May
17 May
19 May
21 May
24 May
Total


Location captured3
Net 2 Net 3
32,30



34 25,27,28
48 35,35
50,60,70,72 35
60,70
31 43,43
11 8



Net 4

27
42
30,47
30,33,35,35,44,46,47



38
12

aNet 2 was located near the mouth  of  Rocky  Run  slough;  net 4 was located in
 the ashpit drain; net 3 was located  in  the main  current  channel draining
 the sedge meadow adjacent to  the  cooling lake  (see  Figure 3).
     No fry were captured in  1977, however.   Following  the  spawning period
that year,  12 fry traps were  placed  in  the marshes  and  were fished from 1
May to 10 June 1977.  Seven traps duplicated  the  1976 effort,  and  five more
were set at other locations in  the marsh.  No fry of northern  pike, walleye,
or muskellunge were caught.   Light trapping on two  nights  in June  1978 also
failed to produce fry.

     Because we caught no pike  fry in 1977, we suspected a poor year-
class.  However, the  1977 year-class was  represented in our 1978 fyke-
netting efforts (refer to Figure  11).   Comparisons  between year-classes
based on the low numbers of fry caught.

Tag Returns—

     During the 3 years of fyke netting,  208  northern pike, 9  muskellunge,
and 39 walleye were tagged on the station site.   To date we have received
three tag returns.  Two northern  pike tagged  in early Spring 1978  were
                                     -25-

-------
caught in May 1978 approximately  17 km downstream by  fishermen  in Lake
Wisconsin near Okee, Wis.  A third northern pike also tagged  in spring 1978
was caught in October 1978, 5 km downstream in the Wisconsin  River.

Effects of Ashpit Effluent on Reproductive Success

     Egg mortalities for the field incubation tests were high at all net
locations (Figure 7).  Eggs incubated in the ashpit drain, however,  had a
higher survival rate (4.6%) after 10 days than eggs incubated in upstream
Rocky Run (3.5%) or downstream Rocky Run (0.2%).  Water quality data in the
ashpit drain showed higher conductivity and warmer temperatures, while total
alkalinity and hardness were reduced (Table 4).  Although more  material
precipitated in sediment traps in the ashpit drain qualitatively it  was very
different from that collected in sediment traps set in Rocky  Run.   The
ashpit drain material was almost entirely the white flocculent  material from
the ashpit, whereas the Rocky Run material consisted  of heavier, brown,
organic sediments that settled on the eggs.

     Newly hatched larvae were held at the same sites as the  egg incubation
tests (Figure 8).  Again, survival was better in the  ashpit drain (54.8%)
compared to downstream Rocky Run  (39.7%) or upstream  Rocky Run  (6.6%).
Although yolk-sac fry survival was better than egg survival,  we still felt
siltation confounded the results and hence turned to  controlled laboratory
bioassays to assess the effect of the ashpit effluent on fish reproduction.

Trace-Element Analysis of Fry—

     Northern pike fry hatched at various locations in the marsh during the
above field incubation tests were analyzed for 20 trace elements (Table
5).  The 10 days that these fish spent in the marsh covered the
embryonically active eyed-egg period through the early larval period.  For
all incubation sites a general increase in trace-element levels was  found in
fry as compared to levels in eggs.  This observation  reflects both embryonic
development and the greater ionic constitution of the water compared to
waters where the parents were caught (Lake Butte des  Morts, Winnebago Co.,
Wis.).  Fry hatched in the ashpit drain contained elevated levels of only
one element, sodium, compared to fry hatched at marsh locations not
influenced by the ashpit effluent.  High iron levels  were found in fry
hatched in upper Rocky Run but the reason for the elevated iron levels is
not known.

Egg Survival in Laboratory Tests—

     In the laboratory, pike egg survival was significantly lower in water
from the ashpit drain than in water from either a control area  in upper
Rocky Run or in a natural mixture of ashpit and Rocky Run water (Figure
9).  Pike eggs were most susceptible to the toxicity  of ashpit  drain water
during the developmental period following gastrulation  (day 3 in Figure
9).  Daily mortality rates were significantly greater for eggs  hatched in
ashpit water only on days 2 through 6 (p < 0.05, analysis of  variance,
Snedecor and Cochran 1967).
                                    -26-

-------
  100
u
8  10
    1.0
z
LLJ
U
                         EGG   SURVIVAL  1977
                                                    .ASH   PIT
                                                    .ROCKY  RUN
                                                    ..MIXTURE
                         95% HATCH
         1     234567
                                     DAYS
       APRIL 16,1977
8     9     10    11
Figure 7.   Survival of northern pike eggs hatched -in situ in the wetlands at
           the Columbia site  during April 1977.  The ashpit site corresponds
           to net 4 and the Rocky Run site to net 1 in Figure 3.  The
           mixture site corresponds to a natural mixture of Rocky Run  and
           ashpit water at net 2 of Figure 3.  Ten bottles with approximately
           50 eggs in each were used at each site.
                                  -27-

-------
             TABLE 4.  WATER QUALITY DATA FOR VARIOUS STATIONS
                    IN THE MARSH—MARCH 21-MAY 12, 1977


Temperature
(°C)


Current
Speed (cm/s)


Dissolved oxygen
(mg/liter)


Conductivity
(ymhos/cm)
@ 25°C

Total alkalinity
(ppm)


Hardness
(ppm)


PH



Sediment
(ml/day)


Turbidity
(JTU)


Upper
Rocky Run
12. la
±5.8
(4.5-20.5)
9
25.6
±5.5
(20.6-36.8)
7
0.6
±6.0
(9.6-12.2)
8
485
135
(381-831)
9
220
35
(155.1-253.0)
8
301
±57
(234-368)
8
7.9
±0.1
(7.7-8.0)
/ 9
1.0
±0.6
(0.5-1.8)
4
36.4
±12.8
(13.7-54)
8
Lower
Rocky Run
13.5
±6.8
(3.9-23.0)
9
19.9
±7.3
(10.1-30.3)
6
9.8
±1.1
(8.2-11.6)
8
870
460
(448-2,028)
9
203
33
149.6-239.8)
8
247
±61
(150-240)
8
8.0
±0.5
(7.7-9.4)
9
0.8
±1.1
(0.1-2.0)
3
31.4
±15.9
(15-57.7)
7
Entrance to
sedge meadow
7.95
±3.8
(3.8-13.0)
4
11.2
±8.0
(2.6-18.3)
3
7.2
±1.6
(5.8-8.9)
3
670
720
(282-1,747)
4
132
9
(126.5-143)
3
167
±16
(150-182)
3
7.8
±0.9
(7.2-9.1)
4
0.8
	
	
1
13.9
±12.1
(0.7-30)
4
Ashpit
drain
14.2
±5.6
(6.2-21.8)
9
19.6
±10.5
(7.4-44.9)
9
9.7
±0.7
(8.3-10.9)
9
1,425
490
(443-1,940)
9
173
33
(133.1-226.6)
8
187
±78
(130-352)
8
7.9
±0.3
(7.4-8.4)
9
2.9
±2.5
(0.5-6.3)
4
19.3
±6.9
(11-34.7)
8

The first entry in each cell is the mean; the second is the standard
deviation; the third is the range; and the fourth is the number of
measurements.
                                  -28-

-------
   100
 ~   50
 u
 CO
 O
 O
     10
 Of.
 Z>
 CO
 o
 of.
    1.0
                                FRY  SURVIVAL
                                       1977
                                                   ,ASH PIT
                                                    ROCKY RUN
                                                   •MIXTURE
                         9596 HATCH
          1     2
        APRIL 16,1977
 6
DAYS
8
9
10    11
Figure 8.   Survival of northern pike larvae placed in wetland at the
           Columbia site for  11 days in April 1977.  The ashpit site
           corresponds to net 4 and the Rocky Run site to net 1 in Figure
           3.  The mixture site corresponds to a natural mixture of Rocky
           Run and ashpit water at net 2 in Figure 3.  Ten bottles with
           50 larvae each were used at each site.
                                 -29-

-------
o
i
               H-
               00
               i-i
               fD
PERCENTAGE  SURVIVAL

                  00
         O        O
                                                                (O
                                                                O
O
O
M 0 H- pd &* W
O fD CO O C C
o PJ o o i-i i-i
3 3 f^ H- <; 3 H-
0<3 O OQ <
OQ Hi rt !3d P)
01 pj C > H1
rt fD 3 *t3
fD pf i-l O
(a fD to n H- Hi
0 C H I—
p4 rt H fD 3
J3* < fD H1 O
j| i-j H- ?^ VO H
(D fD d
O pu CO S! I-1 H-
i-l 3 rt 5^
rt H- 01 fD fD
fD i-l 3 rt i-i
PJ fD i-l fD fD
O PJ Hi (D Cu OQ
3* rt H- PJ OQ
3 M 3 S! CO
rt (D rt PJ
i-l 3 fD 3 rt 3*
fD rt i-l PJ fD PJ
p) CO fD rt H rt
rt . Cu C O
3 H H, 3*
fD C PJ I-i fD
3 H PJ M O &
rt fD rt 3
• 3 tt> 3 H-
i-l H- rt 3
O H- rt fD rt
rt CO C 3"
CO i-l pi fD
X) fD CO
o >-t • y M
Hi fD XJ PJ
CO H- CT1
P) fD *r| rt O
Tl 3 O i-l
*d rt i-l cu PJ
i-l fD i-l rt
O Q- O P) O
X O (-"• i-l
H* PJ 3 3 ^
j>
TO

'
O-
"*•
^ 10
00

CO



4^



/T|



0)


^3 ^J
^^ ^^
*^
CO CIO

(0


—^
^J


^


^^
IO

_ ,
CO





                                                                                        lo
                                                                                        Zoo
                                                                           m 70
                                                                           O
        rt rt
        fD 3"
        t-> fD
                                                                           m
                                                                           >
                                                                           •z.
                                                                                i

-------
           TABLE 5.   CONCENTRATIONS OF TRACE ELEMENTS (ug/g=ppm ON
              FREEZE DRIED WEIGHT  BASIS)  IN NORTHERN  PIKE  EGGS
             AND FRY USED  IN  THE  IN SITU BIDASSAY,  SPRING  1977a
Pike fry hatched at various locations
in the marsh
Element
Br
Sm
La
As
Sb
Na
K
Cr
Se
HE
Ba
Sc
Rb
Eu
Fe
Zn
Co
Ca
Cs
Eggs (before
placement in the marsh)
13+1
0.04+0.01
0.32+0.04
0.49+0.08
0.09+0.01
2,750+27
7,133+277
1.3+0,1
1.6+0.2
0.05+0.01
6.0+1.8
0.07+0.001
6.8+0.7
0.01+0.001
366+16
65+2.0
0.15+0.01
2,533+500
0.04+0.01
Ashpit drain
(n-11)
25+2
0.13+0.01
0.58+0.06
b.d.
0.36+0.06
4,230+50
5,100+1,000
20.0+1.4
9.3+1.1
0.34+0.10
69+20
0.19+0.01
b.d.
0.15+0.03
1,910+100
222+11
0.34+0.07
58,000+8,000
b.d.
Lower
Rocky Run
(n=8)
44+2
0.13+0.02
b.d.
b.d.
b.d.
3,820+70
6,300+1,200
12.8+1.9
5.8+1.4
b.d.
105+34
0.12+0.01
b.d.
b.d.
1,530+160
239+12
0.51+0.10
b
b.d.
Upper
Rocky Run
(n=4)
36+9
b.d.
b.d.
b.d.
b.d.
1,760+120
b.d.
27.0+5.0
7.0+3.0
b.d.
170+50
0.91+0.04
b.d.
b.d.
5,100+400
142+15
1.5+0.3
66,000+3,600
b.d.

aMean = 1 S.D.; b.d. = below detection limits: b = analysis not done.

     None of the filtered-water treatments differed in mortality  (Figure
10); the percentage survival at the end of the experiment was highest in  the
Rocky Run filtered water, but this was not statistically significant.
Whatever factor caused the increased mortality in the ashpit drain was
removed by filtering.  Water quality characteristics differed little in
filtered and unfiltered samples except that turbidity was reduced in
filtered water (Table 6).  Ashpit drain water contained a suspended floe
                                   -31-

-------
     100
      90
      80
   D
   (A
   Ul
   S
   UJ
   o
   oc
   Ul
   Q.
      70
                         EGG   SURVIVAL  1978
                        FILTERED   TREATMENTS
                                      ASH PIT
                                      ROCKY RUN
                                      ,MIXTURE
GASTRULATION
                          95%  HATCH
        t
             1   2  3  4   5   6  7   8   910111213

                                DAYS
          APRIL 6,1978
Figure 10.  Survival of northern pike eggs hatched in the laboratory in
           April 1978 using filtered water from the ashpit drain, Rocky
           Run Creek, and a downstream natural mixture.  Ten lots of
           approximately 100 eggs each were used for each treatment.
                               -32-

-------
           TABLE 6.   WATER CHEMISTRY FOR 1978 LABORATORY EXPERIMENT
           Dissolved                 Total
            oxygen    Conductivity  alkalinity  Hardness            Turbidity
           (mg/liter)   (ymhos/cm  (mg/liter) (mg/liter)     pH      (JTU)
                         at 25°C)

                      Unfiltered Treatments

Upstream      8.9a         390           185         208       8.0       1.0
Rocky Run  (8.3-9.7)    (360-440)    (118-230)   (142-247)  (7.8-8.2)  (0.9-1.8)
               7            7             7776

Ashpit        9.0          1,155          104         158       7.6        .9
           (7.6-9.8)    (310-1,700)  (  99-122)   (141-187)  (7.6-7.7)  (0.6-3.5)
               7            7             7776

Downstream    9.1          525           157         189       8.0       1.2
Rocky Run  (8.5-9.9)    (290-630)    (  89-189)   (184-218)  (7.6-8.2)  (0.6-2.4)
               7            7             7776
                      Filtered Treatments  (3 pm Millepore  filter)

Upstream      9.1          435          184         210        8.0        0.9
Rocky Run  (8.6-9.6)   (375-560)     (115-229)  (144-250)  (7.8-8.2)  (0.7-1.5)
               7             7            7776

Ashpit        9.2          1,200         101         162        7.7        1.3
           (8.1-9.7)   (300-1,540)   (  73-118)   (94-189)  (7.6-7.8)  (0.6-2.3)
               7             7            7776

Downstream    9.0          490          160         189        8.0        1.1
Rocky Run  (7.5-9.7)   (275-595)     (  79-186)  (  99-220)  (7.7-8.2)  (0.7-2.0)
               7             7            7776
aThe first entry in each cell is the mean; the second is  the range; and  the
 third is the number of samples.

that was filtered out and we suspect that this material caused  the
difference in toxicity between filtered and unfiltered ashpit drain water.
Concentrations of trace contaminants in solution were not  thought to  be
altered significantly by filtering because pH, temperature, and conductivity
were similar in both filtered and unfiltered water.  Actual trace-element
concentrations were not determined, however.  Routine water quality
measurements showed ashpit drain water to have a higher conductivity  and
lower pH than water from the other sources (Table  6).
                                   -33-

-------
Year-Class Strengths

     Our objective was to determine if harmful  effects  from construction and
operation of a coal-fired power plant on a pike  spawning  ground could be
detected as weak or missing year-classes.  Having  only  one  preoperational
sampling and no preconstruction sampling proved  a  limitation in attempting
to relate habitat changes caused by the plant  to spawning success of
northern pike.  Our data show typical age-dependent  mortality for northern
pike (Figure 11) with most spawners between  2  and  4  years old.

     The 1971 year-class was the first affected by major  construction on the
generating station site  (Table 7).  Fish from  that year-class were already 5
years old at the time of our first sampling  in 1976.  If  destroying a
significant portion of the spawning marsh  resulted in a loss of some fixed
percentage of pike fry,  fewer fish in all  subsequent years  would reflect
this loss.  Hence, the northern pike spawning  population may be reduced
because of habitat loss, but without historical data on the population
before  1971, this loss would be undetectable.   Since the 1973 year-class was
well represented in our  samples (Figure  11), construction of the ashpit
dikes in 1973 did not result in a detectable reduction  of year class
strength.

     Because the plant began operation  in  mid-spring 1975,  any effects from
the ashpit effluent would not be evident until the 1976 spring spawning
season.  Our 1978 catch  data indicate a  reduced 1976 year-class.  Two-year-
old fish constituted a large portion of  the  1976 (17%)  and  1977 (37%)
catches, but represented only 7% of  the  1978 catch.   Water  levels for the
1976 spawning season were similar to those of  1975 and  1974; therefore,
adequate spawning habitat was available  for  all three year  classes.

     Only in 1977 were enough walleyes captured to allow assessment of
population age  structure (Figure 12).   The 1976 year-class  was only 1 yr old
at that time and hence not very susceptible  to our sampling techniques.  As
this would be the first  year-class to be affected  by the ashpit effluent, we
can not determine what effects the effluent  might  have  on walleye
reproduction.   Furthermore, this species naturally undergoes wide
fluctuations in year-class strengths (Kelso  and Ward 1977).

Importance of this Site  to the Wisconsin River Fishery

     Wisconsin  River wetlands were grouped into 13 major areas of potential
northern pike spawning habitat (Figures  13 and 14, Table 8, Appendix B).
Because no field data exist to show  if northern pike actually utilized these
areas for spawning, the  comparison of spawning areas was accomplished solely
by whether suitable vegetation types existed.   Vegetation types were
determined by infrared aerial photography.  This comparison shows the
wetland area on the plant site (bordered by  Duck Creek on the north, Rocky
Run on  the south, and  the Chicago, Milwaukee,  St.  Raul, and Pacific Railroad
tracks  on the east) to be only a small  percentage  of the total remaining
wetland between the Wisconsin  Dells  and  Prairie du Sac dans (Table  8).  Only
13.1% of the deep-water  sedge meadow and  0.5%  of the shallow-water  sedge
meadow  areas are on the  station  site.   Other areas with substantial deep and

                                     -34-

-------
X
o
6
o
       60
       40
       20
                                                        1978
                                                        n=i23
            69
               70
71
72     73     74
                                                75
       60
    40
       20
    O
    z
    o
DO
E
z
o
o
UJ
    H  60
    UJ
    o
    oc
    S  40
       20
76     77

  1977
  n=59
                  70
                            72     73     74     75
                                 76    77

                                   1976
                                   n=35
                                       75
           69     70     71     72     73    74     75     76     77    78
                                  YEAR CLASS
Figure 11.  Population-age structure of northern pike caught  on the Columbia
            site in 1976, 1977,  and 1978.   For each year  fish caught in all
            nets were pooled to  form one population.   Sample  size  is given
            by n; the cooling lake dikes were built in 1971;  the ashpit
            drain was built in 1973, and plant operation  began in  1975.
                                    -35-

-------
u
s
u
    40
o

z

°  30
GO
O
u
LJJ
O
u

LJU
Q-
     20
     10
            WALLEYE 1977
                  n=34
        1968 1969 1970 1971 1972 1973 1974 1975 1976 1977
              YEAR OF HATCH (YEAR CLASS)
 Figure 12.  Population-age structure of walleye caught on the Columbia site

          in spring 1977. Total sample size was 34 fish.
                            -36-

-------
   PRAIRIE da
    SAC DAM
                           Kilometers
Figure 13.  Major areas  of  potential northern pike spawning  habitat in the
            Wisconsin River and tributaries near the  Columbia Generating
            Station.  Figure 14 shows details.
                                     -37-

-------
 I
u>
00
        KEY:



      WATER


      DEEP WATER SEDGE


I"7"]   SHALLOW WATER SEDGE


|[||jjjj||   CATTAIL


*£:Z   SWAMP-SHRUB


prp   OTHER
            Figure 14a.   Detailed map  of area A (Figure 13).

-------
     o
               KILOMETERS
Figure 14b.  Detailed map of area B (Figure 13).   (See Figure 14a. for key)

-------
-P-
o
I
            Figure 14c.   Detailed map of area C (Figure 13).  (See Figure 14a. for key.)

-------
                  KILOMETERS
Figure 14d.  Detailed map  of  area D (Figure 13).   (See Figure 14a. for key.)

-------
N3
                                                                                       KILOMETERS
                                                                                               \
          Figure 14e.   Detailed map of area E  (Figure 13).  (See Figure  14a.  for key.)

-------
                                                         r          . —4
                KILOMETERS
Figure 14f.  Detailed map  of area F (Figure 13).   (See Figure  14a. for key.)

-------
I
.p-
                                 V,


                                MILES




                                 0
                              KILOMETERS
            Figure 14g.  Detailed map of area G  (Figure 13).   (See Figure  14a.  for key.)

-------
   TABLE  7.   RELATIONSHIP OF  CONSTRUCTION ACTIVITIES TO PIKE YEAR-CLASSES
  Date
            Event
Effect on year-class
January 1971
1973

May 1975
to present
First bulldozer, cooling
lake, dikes completed by April

Ashpit dikes built

Generating station in
operation
  Unknown
None detected

Possibly reduced year-class
for 1976; too early to
detect changes in
subsequent year-classes
shallow sedge-meadow areas are upper Duck  Creek, upper  Rocky  Run  Creek,
Powers Creek and its tributaries, and Lodi Marsh.

     Although we know from our fyke netting  that the  Rocky  Run  A  and  B areas
(described in Appendix B) are not important  spawning  grounds,  they were
included in the analysis to avoid biases.  On  the  basis of  aerial
photography (the criteria used for judging sites where  no fish  survey data
existed) they appeared suitable; therefore,  it would  have been  unfair to
exclude them.  Information provided by  the Wisconsin  Department of Natural
Resources (J. Chizek, personal communication)  indicates that  the  Lodi marsh
is also not used as a pike spawning ground.  If these areas are eliminated
from the comparison, the station site contains 22.1%  of the deep-water sedge
meadow and 0.8% of the shallow-water sedge meadow.  The relative  importance
of the generating station wetlands would probably  increase  even further  if
fish-survey data for other marshes were available  and additional  areas were
eliminated as spawning grounds.

     The largest habitat loss due to construction  of  the generating station
involves replacement of 203 ha of shallow-^water sedge meadow  with the
cooling lake.  This represents a loss of 18% of this  wetland  type formerly
occurring in this section of the Wisconsin River.   Before destruction of
this wetland, the station site contained 28% of the shallow-water sedge
meadow likely to be used by spawning northern  pike  as opposed to  the  current
figure of 0.8%.

DISCUSSION

     Our work has documented the use of wetlands on the site  of the Columbia
Generating Station as spring spawning grounds  for  several important game
fish.  The extent of use varies with annual  flood  levels, but under normal
and high water levels fish enter both the  sedge meadow  adjacent to the
                                    -45-

-------
rr i-j (D >"t C (T» Q. (D P % 7? r^ O O fi i-h &>
(u fu 3 i-( rt re P* re J^* re re 3 ^* T1?* ^*rrrt
i-1 cr rr H. 3" j*r to ft CD P- ^ •< re P*
WOP* 3 wSOwEJ o >•* o
OnflJoiufiJp OO 5O :« <-l 3
re n re ^ »-< loiccron
30 I*" Cfi'jJ'-lCSSreOCn
P* O H O 3" O re fl> ?^ 3 P*
< <-( 3 i-t fi re re CO > CO rt
(D fD PJ i-t p ^* pr rrfD
^ re n> FU 3 p- »-(
?T rr 3 £
? 00 0
C p.
? I
O^ UJ -t* O "vl ro -^j ho ^- Ui (s? CO CO Ln H^ "~*J ^- ro CO LO ^O \O \O Cf^ ^ LO O
LO isJLO ^D*- WLn COOO O>^J LOCO 4^<^ t-^i LnC^ *-Ln t-CO LoC^ ^-LO


^ ^ ho ^ - x-x ro
ON 04> ^-NJ C*-J 0-P- >-^ 4>hO ^Jh- 000 Ui-P- K)tsi 4^K- O>00 O-P-
LO J^-^- J^-CT' -f>CO Ln>^ LOLn Ln-P- ^DhO ^QvO O'-D -P-i— ' LnCO LOhO LH-P'

CO X-N LOLOl—t— 'x-v4^u>
^ i^^ ^^S^i^C^
- X-N O '"^ 1— •
^100 JN'-J )-«Ln (sJhOi— OhOtSJCOOO^-CnW
"-J v^ O (— • -P" "*«J J^« Oo CT^ ho t— •
• •••* •* ••••
? p^
N> O NJ

? "? SJ ^SGoIG^
LO Ln -P* tsi CO t— • O t~O >--
O N5 N3 LH Ln (sj hO
• • * • • • •
O 00 Ln CT< O NJ VO
^ x-s x^sl—COJS H-^
•P- hOLO O\H-H-t— O*^J
• *' •••* • •
<^ OLn COCoOLn hoCO
N3 O tO
Ui O Ln
n
in
3*
(U
CO

Dee p ~wa t e r
sedge



Shallow-
water sedge


Cattail

Swamp- shrub
Canary
grass
Grazed
sedge

Drained
wetlands

Mixed
Floating
macrophytes
deep water
Blue joint




g
0
s>
?5

o
5 ™
S S
Is
• ^
<=K
>s
S "
m Tl
o ^
(-H
IS WETLANDS
RE HECTARES
J-H
§H
ss
^




-------
 cooling  lake  and  the  ashpit drain.  Both of these areas have changed because
 of  the construction and  operation of the Columbia Generating Station.

 Wetland  Water-Level and  Vegetation Change

     Much  of  the  original  sedge  meadow at the site was replaced with a
 cooling  lake, and the remaining  meadow is changing from shallow to deep-
 water marsh (Bedford  1977).  This transition is the result of a fourfold
 increase in ground-water discharge rates since the cooling lake was filled
 in  January 1975.   Prior  to that  time,  water levels rose during spring
 floods,  decreased frequently in  summer to below the soil surface, and rose
 again in autumn.   Water  now stands consistently above the soil surface at
 depths of  from 1  to 60 cm.  The  resulting vegetation change from a community
 of  perennial  sedges to one dominated by annual forbs and emergent aquatic
 species  has serious implications for fish reproduction.  The fall dieback of
 sedges leaves much densely matted intact vegetation that provides excellent
 spawning substrate during  spring floods.  By contrast, the autumn die-off of
 annuals  and emergent  aquatics  provides no such vegetation because stalks
 decompose  and are dissipated before spring floods.  Because these
 vegetational  changes  began in 1975 and accelerated through 1976-77, negative
 effects  on fish reproduction would not have been evident until the 1976 and
 later year-classes.   Since these year-classes are just reaching spawning
 age, monitoring of adult spawners should be continued.

 Effect of_  Ashpit  Effluent  on_ Reproductive Success

     Northern pike utilize, and  may even be attracted to, the ashpit drain
 area because  of higher water temperatures.  This is of interest because the
 drain contains elevated  levels of various trace elements and possibly trace
 organics.  McKim  (1977)  reviewed 56 life-cycle toxLctty tests involving a
 variety  of organic and inorganic chemicals and concluded that the embryo-
 larval and early  juvenile  life stages  of fish are the most sensitive to
 toxicants.  These are  the  stages during which the northern pike studied at
 Columbia have  the  greatest  exposure to the ashpit effluent.  The embryo
 period generally  lasts about 2 weeks (Franklin and Smith 1963) and, since
 juvenile pike do  not begin emigrating  from nursery areas until 10-24 days
 after hatching, these  young fish are exposed to any chemicals in the ashpit
 for a minimum of  4-6 weeks.  We  have observed that some pike may even spend
 their entire  first year  in the Rocky Run area.  In addition to exposure
 through  the water, young pike  may also accumulate toxicants via the food
 chain.   Preferred food items follow a  sequence of microcrustacea, insects,
 and vertebrates (chiefly other pike and tadpoles) with increases in fish
 size (Hunt and Carbine 1951).  These food items accumulate certain trace
 elements,  notably  barium and chromuim  (Schoenfield 1978) from the ashpit.
 They may also contain  harmful  trace organic compounds.

     Considering  the elevated  trace-element levels in ashpit drain water and
 invertebrates  (Magnuson  et  al. 1980),  it is surprising that pike fry hatched
 there did not show higher  element  concentrations  than pike  from unaffected
areas.  The only  elevated  element,  sodium,  is of little toxicological
 importance. The reasons  for  the  higher  iron concentrations  in upstream  Rocky
 Run fry are unknown.   The  small  sample  blomass and inherent analytical

                                     -47-

-------
errors cause differences between incubation sites for the other elements  to
be insignificant.

     The lack of apparent biological accumulation of toxic trace elements  in
pike fry hatched in the ashpit drain does not imply that the ashpit effluent
had no negative long-term effects. Because few pike eggs survived  to
hatching, our sample sizes were small.  Also, the 10-day exposure  period  in
the marsh constitutes only a small portion of the 4-6 weeks that the  fry
minimally spend in nursery areas before emigration.  Furthermore,  our  fry
were still in the yolk sac stage when frozen for analysis; hence,  they had
not yet begun feeding and no uptake via the food had occurred.

     Crayfish caged in this same ashpit-drain system accumulated chromium,
zinc, selenium, and iron over a 2-month period (Harrell  1978, llagnuson et
al.  1980).  Trace-element accumulation was also shown for a variety of
invertebrates living in the drain (Helmke et al.  Unpublished).  These
studies indicate that long-term exposure probably does result in biological
accumulation of trace substances from the ashpit effluent.  Yet the survival
of organisms and the 1976 catch of pike fry in the ashpit drain indicate
that ashpit water is not acutely toxic.  Any long-term negative effects from
the trace-element contaminants or flocculent precipitate entering  the Rocky
Run slough will only be evident by monitoring the northern pike population
structure during the next several years.  Weak or missing year-classes that
cannot be attributed to climatic factors will be evidence for such negative
effects.

Results for 1976—

     Despite the changes in these wetlands, pike continued to use  them for
reproduction.   Successful reproduction in the affected areas was documented
in 1976 by the  capture of pike fry in both the ashpit drain and  the outflow
channel from the sedge meadow adjacent to the cooling lake.  However,  the
chemical and physical changes in wetlands were, in many  cases,  just
beginning by 1976,  1 year after the plant began operation.  Therefore,
successful reproduction in 1976 is no guarantee that reproduction  will
remain unaffected in future years.

Results for 1977 and 1978—

     Despite intensive sampling efforts in 1977, no fry  were caught anywhere
in the Rocky Run slough area.  In 1977, spring water levels were very low
and  resulted in limited amounts of good spawning habitat.  Although some
1977 fry were caught as yearlings in  1978  (Figure  11), evidence  of a  weak
1977 year-class may appear in future  fyke-netting efforts.  Sampling  with
light traps, highly efficient at collecting fish larvae  in other areas,
failed to catch any pike fry in 1978, despite high spring water  levels.
Whether these results represent poor hatching success or simply  inadequate
sampling gear is unknown.  Since documentation of  the natural hatch  proved
unfeasible in the extensive Rocky Run slough area, we utilized  two other
methods for assessing reproductive success:  Egg and fry survival  bioassays
and an analysis of  year-class strengths.


                                    -48-

-------
Egg and  Survival Bioassays

     Our attempts  to  study  hatching  success  by in situ  bioassays were
complicated by factors not  directly  related  to the  ashpit  effluent.   Our
hatching rate was  less than 5% at  all  sites; similar work  has shown success
rates to average 60-70%  for Lake Malar,  Sweden,  and 74% for  Lake George,
Minn.  (Franklin and  Smith  1963).  We  feel that  our poor hatch was the
result of excessive siltation in the incubation  bottles and  probably was
anomalous.  Our results  for larval survival  were better, but again excessive
siltation was a problem.  Since accumulation of  natural sediments was least
in the ashpit drain,  the  greater survival of eggs and fry  there may be
attributed to this factor.

     Results from  the laboratory bioassay indicated that unfiltered ashpit
water was toxic to pike eggs, but  only during the developmental period
following gastrulation.   Survivorship  was not monitored during the early
juvenile period, but  based  on literature reports, we expect  this to also be
a sensitive life stage.   When ashpit water was filtered, the milky white
suspended floe was removed.  Since filtering eliminated the  toxicity of
ashpit water to pike  eggs,  the implication is that  this floe is harmful to
egg survival.  We  aerated the eggs to minimize settling out  of any suspended
materials.  Since  such settling out  might be greater under some field
conditions and since  we have observed  extensive  areas on and near the site
where the marsh bottom is covered  by the floe, detrimental effects on egg
survival are likely.

Analysis of Year-Class Strength

     Analysis of year-class  strengths  revealed no negative effects due to
the construction and  early  operation of  the  generating  station.   ttowever, it
is too early to assess the  eventual  effects  of the  station on pike year-
class strengths.   Although  the station began operating  in  the spring of
1975, physical and chemical  changes  in the area's wetlands were just
beginning to become obvious  in 1976 and  1977.  With the operation of an
additional power generating  unit in  1978, further effects  may be expected.
Our fyke-netting efforts  have provided some  background  data  from which
further changes in pike populations might be detected.   Since it will be
several years at least until the chemical and physical  alterations of the
marsh result in the establishment  of a new equilibrium,  monitoring of adult
year-class strengths  will be an important method of assessing reproductive
success of the game fishes.

Success of Tagging

     The failure to recapture any  fish tagged in previous  spawning runs
could indicate that fyke  nets are  inefficient during  extreme  flood periods
or that northern pike do  not necessarily return  to  the  same  spawning grounds
each year.  During peak flood periods, water  levels  were  occasionally 1  m
over the nets.  Hence, fish could  move upstream  over  the net  or could use
alternate,  temporary  channels.  This hypothesis  is  supported  by  the  capture
of several pike in waters upstream of net 2  (entrance to Rocky Run slough)
that had not been captured in that net.

                                   -49-

-------
     To improve the capture of adult spawners, we suggest  that after  the
upstream spawning run is completed, nets be turned around  to  catch  fish
leaving the marsh after spawning.  Since water levels  are  usually lower
then, net efficiency should be enhanced.  This procedure would also allow an
estimate of the adult spawning population by  the Petersen  tag and recapture
method (Priegel and Krohn 1975).

     The other explanation for the lack of recapture involves the issue  of
whether or not northern pike home to certain  grounds.   Such behavior  is well
documented in salmon and trout (Hasler et al.  1978) and has been suggested
for walleye.  Studies on homing in northern pike are scarce,  but they do  not
suggest such a tendency (Franklin and Smith 1963).  If homing does  occur,
however, then loss of spawning habitat on the  station  site could endanger a
genetically distinct pike population and contribute to the genetic
impoverishment of the species in this portion  of its geographic range.
Sampling of alternate spawning marshes for fish previously tagged on  the
Rocky Run slough area should indicate if pike  use alternate spawning  areas
or return to the same areas yearly.

     The three tags returned by fishermen supported our belief that pike
spend the majority of the year in the Wisconsin River  and  particularly in
Lake Wisconsin.  Therefore, effects from the  power plant are  likely to be
important for only the embryo and early juvenile life  stages, which are  also
the most sensitive to chemical toxicants.  This finding demonstrates  one  of
the many roles flood plains play in the functioning of river  ecosystems  and
how far-reaching the effects of flood-plain disturbance can be.

Inventory of Spawning Areas

     The inventory of potential spawning areas by infrared aerial
photography shows that many alternate sites exist in this  portion of  the
Wisconsin River.  Strict inventory by photo-interpretation can be
misleading, however, since our data show that fish do  not  utilize the upper
stretches of Rocky Run Creek, even though considerable sedge  meadow is
available there.  Inclusion of such areas in  the overall estimate of
spawning habitat would underestimate  the importance of smaller,  but more
productive, wetlands.  Nevertheless, areas currently affected by the
Columbia Generating Station do not constitute the major portion  of  the
wetland areas remaining in this section of the Wisconsin River.  Wetlands on
the plant site are known to attract large numbers of spawning fish, however,
and therefore their importance to the Wisconsin River  fishery should  not  be
overlooked.  Future degradation of the area wetlands should be minimized or
else the loss of further documented spawning  marsh is  likely.

Final Considerations

     We  feel that our study has successfully  documented the use  of  areas
affected by the power plant as pike spawning  grounds.  In situ egg hatching
bioassays were inconclusive, but  our  laboratory bioassays  indicate  that
ashpit-drain water is toxic to certain developmental stages of pike
embryos.  One of the major considerations  in  determining  the  importance of
this area to the Wisconsin River  fishery is whether unique homing  fish

                                     -50-

-------
stocks are involved.  In addition to northern pike, muskellunge, and
walleye, other spring spawning fish, such as catostomids, are affected.   If
homing is important, then loss of habitat involves loss of these stocks.   If
fish simply search for the best suitable habitat, then loss of wetlands on
Rocky Run Creek might be compensated for by a switch-over to alternate
spawning sites, but this could also represent a reduced area for young-of-
the-year production.

     Finally, the year-classes of 1976 and later should be studied  carefully
because they will probably reflect the long-term effect of the plant.  A
series of weak year-classes that cannot be correlated with low water  levels
or extreme water-temperature changes, and which are not evident in  spawning
populations at other sites, would be a strong indication of reproductive
failure due to operation of the Columbia Generating Station.
                                     -51-

-------
                                  SECTION 4

                ZINC  TOLERANCE IN FOUR GENERATIONS OF FLAGFISH
INTRODUCTION

     In Section 2 we observed that, despite  the  changes  in wetland water
levels and substrate, northern pike and other game  fish  continued  to  use the
affected marsh area at the Columbia site  for spawning.   In this  section we
will examine a question that may arise at any wetland site affected by the
construction and operation of a generating station:  Can fish populations
adapt genetically to increased trace-element levels  in the environment by
evolving tolerance?  Tolerance is defined as the relative capacity of an
organism to grow or thrive when subjected to a normally  unfavorable
environmental factor.  Chronic exposure to toxicants may favor those
individuals with a genetic make-up that confers  resistance.   Through the
process of natural selection, the tolerant members  of the population survive
and transmit the trait of tolerance to their offspring.

LITERATURE REVIEW OF METAL TOLERANCE

     Contaminants can cause selective pressures  that result in tolerant
populations as shown by the many documented  cases of pesticide-resistant
insects (Crow 1957).  The number of insect species  and the types of
chemicals involved are numerous, but two  general observations are
noteworthy:  (1) the evolution of tolerance  is rapid and (2)  the mechanisms
of resistance are diverse.  Among the mechanisms identified are  the
development of behavior patterns that lessen exposure to the poison,
decreased uptake through reduced permeability of the cuticle, and  enzymatic
detoxification.

     The widespread and often indiscriminate use of chemical poisons has
subjected many other types of organisms to similar  selective pressures.
Studies have found pesticide-resistant populations  of fish (Vinson et al.
1963), crayfish (Albaugh 1972), frogs (Ferguson  and Gilbert 1967), and mice
(Webb and Horsthall  1967).  In many cases resistance may be an acquired
trait through inducible detoxifying enzymes  (Webb and Horsthall  1967, Brown
1976), but it can also be inherent in certain organisms  regardless of prior
pesticide exposure (Crow 1957, Ferguson 1967).

     In addition to synthetic organic chemicals, tolerance to trace-element
contaminants also has been documented.  Studies  of  both  terrestrial plants
(Antonovics et al. 1971) and aquatic plants  (Stokes et al. 1973, McLean and
Jones 1975) show again that the evolution of tolerance is rapid, involves


                                      -52-

-------
diverse mechanisms, and can occur in response  to a variety  of  trace
elements.

     Tolerance to metals has been demonstrated in recent  studies  of
invertebrates from contaminated aquatic habitats.  Bryan  (1976) found  the
marine polychaete Nereis diversicolor  to have  zinc,  copper,  silver,  and
possibly lead-tolerant populations.  Tolerance was limited  generally to the
metals in a particular habitat.  Laboratory  experiments indicated the
mechanism was based on reduced permeability  to metal  ions.   Whether
tolerance was an acquired  (inducible)  or an  inherited trait  was not
determined.  B. Brown (1976) found copper and  lead-resistant populations  of
Asellus meridianus Rac. in rivers with a history of  metal pollution  from
abandoned mines.  Both acute lethal bioassays  and chronic growth  studies
demonstrated tolerance.  The persistence of  tolerance in  second generation
organisms from a laboratory culture indicated  a genetic basis  for this
trait.

     Despite such demonstrations of the evolution of  metal  tolerance in
aquatic plants and invertebrates, no examples  of naturally metal-tolerant
fish populations have been reported in the literature.  Although  waters with
high metal contamination frequently contain  plants and invertebrates,  fish
are conspicuously absent (Carpenter 1924, Jones 1958,  Weatherly et al.
1967).  In waters with slightly elevated metal levels, fish  populations
exist but reproduction is  severely depressed and their long-term  survival is
uncertain (Van Loon and Beamish 1977,  McFarlane and  Franzin 1978).

OBJECTIVE OF THIS STUDY

     This study tested the hypothesis  that if  genetic factors  partially
determine the variation in susceptibility to metal toxicants,  the resistance
of a fish population can be increased  through  selection.  The  ability  of
fish to develop tolerance is of interest since  trace  elements  are often
released by coal-fired power plants at levels  that may have  severe long-term
effects on fish populations even though they are not  acutely lethal  (Cherry
and Guthrie 1977).  The approach used  was to breed the flagfish (Jordanella
floridae), a species well suited for extended  laboratory  studies  (Smith
1973), for resistance to the long-term effects  of the element  zinc.

METHODS

Zinc as an Experimental Toxicant

     Zinc was chosen as the toxicant for this  experiment  because  it  is a
common pollutant whose biological effects are well documented  (Skidmore
1964, European Inland Fisheries Advisory Commission  1974).   Use of zinc and
flagfish provided a convenient laboratory model for the selection processes
that would affect other fish species exposed to other trace-element
contaminants.  Fish exposed to zinc may die  from either acute  or  long-term
effects.  Acute mortality,  the result  of extensive gill damage, occurs
within the first 4 days of  exposure to high  zinc concentrations.   Death from
long-term effects occurs at lower concentrations and  involves  damage to
internal organs. To determine the level of zinc resulting in chronic

                                    -53-

-------
mortality, groups of flagfish were exposed to various concentrations  of  the
element (Figure 15).  A concentration of 0.8 mg/liter was used for the
parental generation, but this was increased to 1.0 mg/liter for successive
generations to accelerate mortality and increase the selective pressures
affecting the population.

     To begin the experiment, a stock supply of flagfish was randomly
divided into two groups.  One group (the selected population) underwent
three generations of selection for zinc resistance while the other group
(the unselected population) was maintained under the same laboratory
conditions but received no zinc exposure (Figure 16).  Selection was
accomplished by exposing adult fish to a chronically lethal zinc
concentration until 50 to 60% of the population had died.  To determine  if
selection was increasing zinc resistance, a subsample of each generation of
the unselected population was exposed to zinc along with the selected
population and survival times were compared between the two groups.

Biological Procedures

     The original stock of flagfish was purchased from a commercial supplier
(Ross Socoloff Farms, Bradenton, Fla.) and maintained in tap water in
Madison, Wis.

     A constant photoperiod of 16:8 h (light:dark) and constant temperature
of 25°C were used throughout the experiment.  Fish ate frozen brine shrimp
supplemented with commercial fish food  (Tetra Conditioning Food, Tetra Werke
Co.).  Because of the hardness of Madison tap water (Table 9), zinc
exposures were at a dilution ratio of 1:3 (tap to distilled water).   Fish
were acclimated to the dilution water for 1 week before exposure to zinc.
During the bioassay deaths were recorded at 12-h intervals; death was
defined as the failure to respond to a mechanical stimulus.
     Breeding proceeded according to methods outlined by Spehar (1976).
Breeding tanks were 55-liter aquaria divided into thirds by plexiglass
partitions.  Eggs, deposited on orlon yarn spawning substrates, were
collected daily.  Eggs hatched in egg cups made of PVC piping cut to  6-cm
lengths and covered with plastic screen at one end.  Larvae ate newly
hatched brine shrimp nauplii until old enough to take frozen adult brine
shrimp.  Generation time was approximately 6 months.

Bioassay Procedure

     The flow-through bioassays were conducted along guidelines issued by
the U.S. Environmental  Protection Agency (1975).  The characteristics of  the
tap and dilution water are given in Table 9.  Zinc concentrations,
conductivity, and temperature were monitored daily; hardness, total
alkalinity, pH, and dissolved oxygen were measured weekly.  All measurements
were taken according to methods outlined by American  Public Health
Association et al. (1975).  Zinc samples, some unfiltered and some filtered
through 0.4 um nucleopore filters, were measured by atomic absorption
spectrophotometry.  Filtering allows an estimate of how much zinc is
actually in solution and hence readily available for uptake by fish  (U.S.
Environmental Protection Agency  1975).  Nominal  (added) zinc concentrations,

                                    -54-

-------
I
Ui
Ul
I
                     90
                  t  50
                  QC
                  O
HI
O
<

z
HI
O
DC
UI
Q.
            10
                         0.5
                                                    TIME   [DAYS]
                                                                            10
                                                                                20
Figure 15.
                     Cumulative mortality  (probit scale) as a function of exposure time for flagfish exposed

                     to three  zinc  concentrations.
                     D 5 mg/liter
                                      • 2.5  mg/liter
                                                         00.8 mg/liter

-------
                      STOCK POPULATION
                                                                  GENERATION
UNSELECTED POPULATION
     SELECTED  POPULATIOK
  b
<=>
                                                   ZN EXPOSURE
                                                   OFFSPRING
                                                  ZN EXPOSURE
                                                  OFFSPRING
                                                  ZN EXPOSURE
                                                  OFFSPRING
                                                  ZN EXPOSURE
                                                                   PARENTAL
Figure 16.  Procedure used in selecting for zinc resistance in laboratory
            populations of flagfish.  An initial stock population was
            randomly divided into two groups; one underwent three genera-
            tions of selection for zinc resistance while the other remained
            as a control population.

           aSurvivors bred to obtain data on spawning success but larvae
            were then discarded.
           ^Comparisons of survival times between these groups were used
            to assess effects of selection.
                                   -56-

-------
length of exposure, and recovery time before breeding for each generation
are shown in Table 10.

        TABLE 9.   CHEMICAL  CHARACTERISTICS  OF MADISON, WIS.,  TAP WATER
              IN WHICH FLAGFISH WERE RAISED, AND DILUTION WATER
                    IN WHICH ZINC  EXPOSURES WERE CONDUCTED
 Item
Tap water
 Dilution water
(weekly samples,
all generations)
Hardness
(mg/liter CaC03)

Total alkalinity
(mg/liter CaC03)

Conductivity
(umhos/cm at 25°C)

PH
Dissolved
oxygen (mg/liter)

Zinc (ug/liter)
   28 Oa
(235-296)
   590
(510-640)

   7.4
(7.2-7.8)
    20
 (10-50)
       73
    (61-88)

       59
    (44-67)

      140
   (120-180)

      7.3
   (6.7-7.9)

      7.6
   (6.7-8.1)

       23
     (2-45)
lFirst entry in each cell is the mean and the second is the range.
'Not measured.
            TABLE  10.  NOMINAL  ZINC  CONCENTRATION  (ppm),  LENGTH OF
       EXPOSURE (DAYS),  AND RECOVERY TIME BEFORE BREEDING (WEEKS), FOR
    THE  ZINC  EXPOSURES  OF  THE  PARENTAL  AND  THREE  GENERATIONS OF FLAGFISH

Nominal zinc
Length of
concentration exposure
Generation
Parental
First
Second
Third
(ppm)
0.8
1.0
1.0
1.0
(days)
17
17
6.5
13
Selection intensity
(% dead for selected
population)
60.0
53.7
49.3
55.4
Recovery time
before breeding
(weeks )
8
3
2
1
                                    -57-

-------
RESULTS—EXPOSURES AND CALCULATIONS
Parental Generation
     Parental generation flagfish of the selected population were exposed to
0.8 mg/liter zinc for a 17-day period until 60% had died (Figure 17A).
Unselected population fish were not exposed to zinc.  Approximately 75% of
the zinc (Table 11) was in a soluble form considered most toxic to fish
(European Inland Fisheries Advisory Commission 1974).

               TABLE  11.   ZINC  CONCENTRATIONS  FOR THE  PARENTAL,
                FIRST, SECOND, AND THIRD GENERATION EXPOSURES
Generation
Zn-exposed fish
  (yg/liter)
    Control fish
     (jig/liter)
              Filtered
        Unfiltered
Filtered
Unfiltered
Parental
First
Second
Third
575+100a
(18)
750+65
(40)
1,122+97
18)
850+76
(18)
811+48
(7)
1,000+140
(11)
1,348
(1)
915+9
(2)
25+1
(2)
22+12
(12)
30+20
(2)
< 50
(3)
40+16
(2)
28+16
(67
40+10
(27
< 50
(1)

aFirst entry in each cell is the mean + standard deviation for water samples
 taken during the course of the zinc exposure; the second is the number  of
 samples.

     Survival times of fish exposed to zinc may be correlated positively
with fish size (Bengtsson 1974).  Therefore, standard lengths of survivors
and nonsurvivors of the initial exposure were compared to determine if large
size and not an inherent resistance to zinc could explain survival (Table
12).  The comparison was based on a two-sample t-test with unequal
variances.  Males that died and those that survived did not differ in length
(t = 1.12, d.f. = 85, p = 0.27) (Table 12).  Females, however, were longer
than nonsurviving females (t = 2.38, d.f. =  1.34, p = 0.02), but the length
difference was small.  The 95% confidence limits for the length difference
shows that surviving females were only 0.3-3.4 mm (0.8-9.9%) longer than
nonsurviving females.  For fish that died during exposure to zinc, there was
no correlation between body length and time to death (r = -0.06 for males
and r = -0.11 for females).  Thus, for the size range of fish used,
                                    -58-

-------
o




80


60

40


20
1

80



60




40


20
• A
i
*
"~ •

• m
- •
"\
'•


_
i i i i i in
1 5 10 15 20 25
* * * 2 o B
. ft 8.

•
o 0*
^ • —
o •

O A
o
•» O ™"
OQ
o
_ —
IVA.


80

LLJ
< 60
LU
O
5 40
LU
O_
20
"
• inn
IUU
80

II i

< 60
LU
3
| 	
Z
,m, 40
^
LLJ
CL.
20
'
o • C
O ^
O
- 2 0 —
o •

* —
0 •
o ~
o

— —
T , , , ff
1 5 10 15 20 25
• • Q D
* 8 o o
- FS °
• "~
• °
0
- *o —

• 0

9
« ^


M
, , , 2
I 5 10 15 20 25 , 5 10 15 20 25
                     TIME (DAYS)
        •Population selected for zinc tolerance.
        oPopulation not selected for zinc tolerance.
TIME (DAYS)
 Figure 17.   Survival rates of exposed  (selected) and  control  (unselected)  flagfish populations over
              four generations.  Survivorship is plotted  against  elapsed  time in days (log scale).   A,
              parental generation; B, first generation  of offspring;  C, second generation; D,  third
              generation.

-------
selection was not for large size but for some other factor that allowed
certain fish to be more resistant to zinc than others.

          TABLE  12.   STANDARD  LENGTHS  (MILLIMETERS)  OF SURVIVORS  AND
          NONSURVIVORS  FOR  PARENTAL-GENERATION FLAGFISH  EXPOSED  TO
                        0.8 mg/LITER ZINC FOR 17 DAYS
 Sex
Nonsurvivors
Survivors

Male


Female

Mean
(S.D.)
n
Mean
(S.D.)
n
36.9
(5.3)
106
34.2
(4.2)
83
38.1
(6.7)
53
36.0
(5.4)
72

     Residual effects on reproduction of a single sustained zinc exposure
were determined by comparing spawning data for both experimental and control
fish.  Since fecundity is affected by fish size, wet weight and standard
lengths of fish used for spawning from both experimental and control
parental populations were compared with a two-sample t-test^with unequal
variances (Table 13).

     TABLE  13.   STANDARD LENGTHS  AND WET WEIGHTS OF PARENTAL  GENERATION
                FLAGFISH USED TO PRODUCE THE FIRST GENERATION

Male
Item length(mm)
Experimental 44a
population (6)
(zinc -exposed) 30
Control ^ 47
population (6)
(not zinc-exposed) 27
Female
length(mm)
42
(4)
30
43
(5)
27
Male wet
weight (g)
3.06
(1.13)
30
3.48
(0.99)
27
Female wet
weight (g)
2.43
(0.58)
30
2.53
(0.98)
27

aThe first entry of each cell is the mean; the second is standard deviation;
 and third is number of fish.

     No difference in weight or length was found for either males or  females
(p > 0.5).  Size differences therefore were not responsible for any
fecundity differences.  The spawning data, not normally distributed,  were
analyzed by the Mann-Whitney U test (Siegel 1956).  The two populations did
not differ in the number of days until the first eggs were laid  (p =  1.00)
or in the percentage hatch of eggs  (p = 0.80) (Table 14).  Experimental fish
                                    -60-

-------
averaged more eggs per spawn than control fish  (p =  0.03) with a  95%
confidence limit of 0.4 to 13.1 more eggs per spawn.  Thus, fish  surviving  a
concentration of zinc lethal to the majority of the  population showed no
harmful residual effects on reproduction after  8 weeks recovery time.

          TABLE  14.   SPAWNING DATA FOR PARENTAL  GENERATION FLAGFISH3

Item
Experimental
population
(zinc-exposed)
Control
population
(not zinc-exposed)
Days until
first spawn
9b
(5-20)
20
8
(15-17)
13
Number of
eggs/spawn
8.2
(0-86.7)
21
7.6
(0-53.6)
15
Percentage
hatch
74.7
(0-100)
19
76.6
(0.4-96.4)
13

Experimental fish survived a  17-day exposure to 0.8 mg/liter  zinc.   Control
 fish were not exposed to zinc.
 First entry is median; second is range; and third is  number  of  fish pairs.

First Generation

     The effect of one generation of selection for zinc  resistance was
determined by comparing the survival of first-generation fish  from both  the
selected and unselected populations exposed to 1 mg/liter  zinc.   As  in the
initial exposure, about 75% of the zinc was in a soluble form  (Table 11).
     The first deaths from zinc exposure occurred in the first-generation
(Fj) selected population (Figure 17B).  Once fish began  to die in the zinc-
exposed unselected population, however, mortality was higher and  surpassed
the experimental population after 8 days.  After 17 days,  68.9% of the
unselected population but only 53.7% of the selected population had  died.
Median survival times were 12 days for the zinc-exposed  unselected
population and 16.5 days for the selected population.  A Mantel-Haenszel
test was used to determine the significance of this difference (Snedecor and
Cochran 1967).  This test computes x  values for the observed  deaths during
each time period, given the number of fish alive at the  start  of  the period
and the null hypothesis that there is no difference in susceptibility to
zinc poisoning between the groups.  Each time period is  treated
independently, and the dependency on previous events (as with  a parameter
such as cumulative mortality) is avoided.  Survival times were significantly
different (x2 = 9.9, p = 0.001), indicating that selected  population fish
were less suceptible to zinc poisoning than unselected population fish.
     A comparison of standard lengths of F, selected and  zinc-exposed
control unselected indicated that size differences were not a factor in  the
increased resistance of the experimental group  (t-test, p = 0.85).
                                    -61-

-------
     No significant differences in length or weight for either sex were
found that might affect reproduction (Table 15) (t-test, p > 0.05).  Neither
did the groups differ significantly in time to first spawn, average number
of eggs per spawn, or the percentage of hatched eggs (Table 16)  (Kruskal-
Wallis Test, p > 0.05, Siegel 1956).

                   TABLE  15.  LENGTHS AND WET  WEIGHTS  OF
                   FIRST-GENERATION FISH USED FOR SPAWNING



Item
Selected
population
(zinc-exposed)
Unselected
population
(zinc -exposed)
Unselected
population
(control)

Male
length (mm)
34a
(6)
12
37
(6)
8
33
(7)
14

Female
length (mm)
30
(3)
12
33
(2)
8
32
(4)
14
Male
wet
weight (g)
1.24
(0.62)
12
1.53
(0.57)
8
1.24
(0.81)
14
Female
wet
weight(g)
0.89
(0.28)
12
1.12
(0.18)
8
1.05
(0.37)
14

aThe first entry is the mean; second is standard deviation; and  the  third  is
 number of spawning pairs.

Second Generation

     The effect of breeding two generations for resistance to  zinc was
determined by comparing fish survival from both selected and unselected
populations when exposed to a lethal zinc level (Figure  17C).  Zinc  analysis
(Table 11) revealed a slightly higher zinc concentration (1.1  mg/liter)  than
the intended 1.0-mg/liter level.

     As in the first generation, second-generation fish  from the selected
population proved more resistant to zinc poisoning than  fish from the
unselected population.  After zinc exposure of 6.5 days, mortality was 68.5%
for the zinc-exposed unselected population, but only  49.3% for the selected
population.  Median survival times were 5.25 and 6.4  days for  the
zinc-exposed control and experimental populations, respectively.  The
difference in survivorship was significant at the p = 0.004 level (x1  =8.55,
Mantel-Haenszel test).
1
     Selected-population  fish were  significantly  longer  (p  = 0.05)  than
unselected-population fish with average  standard  lengths  of 23.8 and  25.5
mm, respectively.   Considering results  from  the parental  generation,  the
size difference probably  was not a  significant source  of  variability  in
zinc-tolerance levels between these two  populations.   The two groups  did  not
                                    -62-

-------
    TABLE 16.  SUMMARY OF SPAWNING DATA FOR Fj-F3 GENERATIONS OF FLAGFISH
Generation
F,
1

Unselected pop-control


Unselected pop — Zn exposed


Selected pop — Zn exposed

F2

Unselected pop-control


Selected pop — Zn exposed

F.,
3

Unselected pop-control


Unselected pop — Zn exposed


Selected pop — Zn exposed

Days until
first spawn3


11
(6-18)c
4
7
(5-20)
8
6
(3-21)
7

4
(4-6)
7
7
(4-14)
6


33
(23-37)
10
25
(18-31)
8
22
(18-25)
4
No. of
eggs/spawn


0
(0-6.1)
10
2.2
(0-3.5)
10
3.7
(0-13)
9

2.4
(0-27.6)
11
0
(0-6.7)
14


0
(0-12.7)
10
2.0
(0-18.3)
11
0
(0-8.9)
12
Hatch (%)a


89
(2-100)
4
90
(50-100)
7
81
(4-100)
7

84.8
(73.7-100)
7
73.7
(0-86.4)



8.7
(83-94)
4
99
(89-100)
7
92
(71-100)
4

 Only spawners producing eggs are included.
 Includes all spawners with pairs producing no eggs counted as zero.
GThe first entry is the median, the second is the range, and the third is
 number of fish pairs.
differ in the number of days to first spawn (p = 0.15), the average number
of eggs per spawn (p = 0.22), or the percentage of hatched eggs (p = 0.1)
(Mann-Whitney U Test) (Table 16).
                                   -63-

-------
Third Generation

     The effect of breeding three generations  for  resistance  to zinc was
determined by comparing survival of fish from  the  experimental  and  zinc-
exposed control populations when exposed to a  lethal  zinc  level (Figure
17D).  Zinc water analyses are given in Table  11.

     In contrast with the first two generations, mortality was  higher in
third-generation selected-population fish  than in  third-generation
unselected-population fish.  After an exposure period of  13 days,  46.8% of
the zinc-exposed unselected population and 55.4% of the selected-population
fish had died.  Median survival times were 13.5 days  for  the  zinc-exposed
unselected population and 12.5 days for the selected  population.   Despite
this decline in tolerance for third-generation experimental ftsh,  the
difference in survival between the two groups  was  not statistically
significant (xf = 1.63,  p = 0.20, Mantel-Haenszel test).   The  two
populations did not differ in standard lengths (t-test, p  = 0.59);  hence,
size was not a factor in the failure of the experimental population to
display superior tolerance.

     The only difference in spawning success between  selected nad  unselected
population fish exposed to zinc and unselected population  not exposed to
zinc was that the latter group took longer to  produce eggs than the other
groups (Kruskal-Wallis Test, xf = H.l.  P = 0.05) (Table  16).

     The results of breeding three generations of  flagfish for  resistance to
zinc are summarized in Figure 18.  The absence of  a  trend  toward  increased
tolerance with continued selection suggests that  two  confounding  factors may
be involved, inbreeding depression and carry-over  effects.  Both  phenomena
are discussed below.

DISCUSSION

Variability ^Ln Zinc Tolerance

     These experiments demonstrate considerable variation  in the  tolerance
of fish to zinc.  In  the parental generation,  for  instance, some  fish died
after only several days of exposure to zinc whereas  others lived  for the
entire 17-day period.  In a similar study  (Bengtsson  1974) some fish
survived a 100-day exposure to concentrations  of  zinc that were lethal to
the majority of the population.

     As with lethal tests, sublethal studies also  reveal  considerable
variation in the response of organisms to  toxicants.   The  sublethal effects
of contaminants are being examined in many bioassays  ranging from the
subcellular to the community level (McKim  et al.  1976, Sprague  1971, G.W.
Brown 1976, Maki and  Johnson 1976).  In a  study of the sublethal  effects of
copper on the coujh frequency, locomotor activity, and feeding  behavior of
brook trout, response variation was evident  (Drummond et  al.  1973).  The
cough frequency of some fish increased markedly,  whereas  that of  others did
not change.
                                     -64-

-------
Ui

I
                  -~
                  40
          u
          LJU  ±-

          U-  >
          u.
                  20
LU

O Z    0

£ <

z a
LLJ LU
          Lti     -20
          Q_
                                1234


                                    GENERATIONS OF SELECTION
      Figure 18.  Summary of three generations of selection for zinc resistance in flagfish.

-------
     The two sources of this variation, environmental and genetic  factors,
must be understood if we are to predict the long-term effects of
contaminants.  Environmental factors alone can have a dramatic effect  on an
organism's response.  In a study by Spehar (1976) flagfish larvae  exposed as
embryos to zinc and cadmium were more tolerant than unexposed larvae,
although both sets were produced by unexposed parents.

     Differences in the response of organisms to a pollutant are termed
"phenotypic variability" (observable variability due to both an organism's
heredity and environment).  Even in bioassays with strains of laboratory
fish raised under uniform conditions, much phenotypic variability
persists.  Such variability might be caused by differences in the  general
health or vigor of individual fish and not by genetic factors conferring
resistance to a particular toxicant.  A study by Sparks et al. (1972)
demonstrates the importance of stress on fish tolerance.  When pairs of
bluegills were exposed to lethal zinc levels in bare aquaria, dominant fish
had a significantly longer survival time. When the experiment was  repeated
with shelters provided, survival times did not differ.  Presumably,
subordinate fish were no longer subject to the additional stress of
harassment by dominant fish.

     If the response variation to toxicants is caused by nongenetic
differences in health and vigor of fish, then toxicants should act simply as
a general stress on the population, culling out the weak individuals.  An
increase in the gene combinations conferring resistance should not
continually increase the tolerance of the population.  Simply stated,
selection should not continually increase the resistance of future
generations.  However, an increase in resistance through successive
generations has been demonstrated for many contaminants and many species
(see Introduction), indicating that genetic factors must be partly
responsible for phenotypic variability.

     Differences in the reaction of fish to toxicants also might be  caused
by factors such as size, sex, age, diet, and degree of acclimation.  In this
study, as in many others, diet and degree of acclimation were carefully
controlled.  We found that size had little effect on survival time for the
parental generation.

     The literature provides conflicting results about the effect  of age on
resistance to a toxicant.  Age was not considered in this study except that
only adult fish were exposed to zinc. Jones (1938) found no difference in
survival times between juvenile and sexually mature sticklebacks exposed to
a range of zinc concentrations.  Bengtsson (1974) reported, however, that
the resistance of the minnow Phoxinus phoxinus to zinc increased with  age.
Adelman et al. (1976) found that even for fish of a constant age or  size
class raised under identical laboratory conditions, variability in response
to a toxicant still exists.

     The presence of this variation in bioassays, despite uniform  laboratory
conditions, suggests that genetic differences between individuals  may  be a
significant source of variability.  Although genetic studies on zinc and
fish have not been done, work on lead and mercury suggests that considerable

                                    -66-

-------
variability  in  response  to metal  pollutants  may  be  genetically based.
Burger  (1973) studied  the inheritance  of  lead  resistance  in  guppies  and
obtained a heritability  estimate  of  0.26  to  0.57 for  survival  time  of  fish
exposed to acutely  lethal levels  of  lead.  [Heritability measures  that
fraction of  the  total  phenotypic  variability due to the additive  effects  of
genes;  this  part of the  phenotypic variability is the most responsive  to
selection  (Crow  and Kimura 1970)].   In a  study of mercury resistance in
steelhead trout, Blanc (1973) estimated the  heritability  to  be 0.5  for
survival time in chronically lethal  levels of  methylmercury.

     Despite these  demonstrations of genetically based differences  in  the
tolerance of fish to toxic metals, we  should be  cautious  of  ascribing  the
variability  found in bioassays  to genetic  sources.  In a  study by Rachlin
and  Perlmutter  (1968)  guppies still  varied in  their response to zinc after
31 generations of inbreeding.   Since such  prolonged inbreeding should  have
greatly reduced  genetic  variability, environmental  factors appear as the
chief source of  the variation,  although the  fish were raised and  exposed
under uniform laboratory conditions.

Effects of Zinc  on  Reproduction

     This study  found  that after  only  2 to 8 weeks  recovery  time, flagfish
surviving a  zinc exposure level lethal to  the  majority of  the  population
were able to reproduce as successfully as  the  control population.  Similar
work with zebrafish indicated that a 9-day period of  zinc exposure  reduced
egg production and  egg fertility, but  these  processes returned to normal
levels within several weeks of  a  return to uncontaminated water (Speranza et
al.  1977). Results  of  these relatively short-term exposures  to zinc  contrast
with results of  exposures encompassing the entire life cycle of fish.   In
the long-term studies, reproduction was reduced  significantly  although
survival was unaffected  (Brungs 1969,  Bengtsson  1974).  Since  contaminants
in an aquatic environment may be  present only  intermittently or temporarily
(Cairns et al.  1971, Leland et  al. 1976), it is  important to know if fish
have the ability to recover from  temporary exposures.  Our work suggests
that such recovery  is possible  for fish exposed  to  zinc.

     Whether selection for resistance  to lethal  effects also causes
resistance to sublethal  effects could  not be determined in this study
because the  zinc-exposed control and experimental populations  for both the
first and second generations did not differ  in spawning success and  no other
sublethal aspects were investigated.   Such a phenomenon might  be  expected in
fish, however, since it has been demonstrated  in metal-tolerant
invertebrates (B.E. Brown 1976).

Selection for Resistance

     After three generations of selection, the selected line was  more
tolerant for the first two generations, but showed no difference  from  the
unselected line in the third generation.  The absence of any trend toward
increased resistance with continued selection  (Figure 18) does  not
necessarily indicate an inability to evolve metal tolerance.   The actual
decline in the relative tolerance of the experimental population  suggests

                                   -67-

-------
that confounding factors are responsible.  Even if none of the variability
in metal tolerance was genetic, the experimental population should have
remained equal to the control line.

     The continued decline in performance of the experimental population
suggests the phenomenon of inbreeding depression.  Although both lines had a
similar number of parents for each generation (Table 17), it has been shown
that in populations undergoing selection, selected parents are related more
closely than randomly chosen parents (Robertson 1961).  Because close
relatives are more likely to produce offspring with harmful homozygous
recessive gene combinations, the result of such inbreeding is a general
depression of survival and vigor in the selected population (Kincaid
1976).  In our experiments, parents of the selected population were chosen
on the basis of having survived an exposure to zinc, whereas unselected
population parents were chosen randomly.  Fish surviving the zinc exposure
were more likely to be closely related, if survival is genetically based,
having inherited favorable genes from a common ancestor. Therefore, the
decline in overall health and vigor because of inbreeding would be greater
for selected-population fish than for zinc-exposed unselected-population
fish.  Hence performance of the experimental population would decline with
increasing generations of selection.

     TABLE  17.   NUMBER OF  PARENTS  CONTRIBUTING LARVAE FOR EACH GENERATION
                      Unselected line                  Selected  line
Generation             Male   Female                   Male    Female
Parental
First
Second
Third
14
16
10
7
11
16
10
7
20
17
11
8
20
17
9
8

     Another possible explanation for the decline of the experimental
population is the carry-over of harmful effects from mother  to  offspring
through the egg cytoplasm.  Exposed fish are expected gradually to  lose zinc
retained in their body tissue upon returning to clean water.  It is possible
that females incorporate some of the zinc into the cytoplasm  of their eggs
and that offspring of females exposed to zinc begin life with elevated zinc
levels.  According to this hypothesis, exposure of these offspring  to zinc
later in life results in increased susceptibility (McKim  1977).  The
magnitude of this carry-over effect would increase with a  decrease  in the
recovery time allotted to females before breeding.  Because  recovery time
was shortened with each generation of selection (Table  10),  any carry-over
effects work against the direction of selection.  By the  third  generation
the parents had had only 2 weeks' recovery  time before  breeding, and the
carry-over effect may have finally negated  the effects  of  selection.


                                   -68-

-------
     The relative importance of inbreeding depression and carry-over effects
in producing the observed decline in zinc tolerance will be tested in future
work.  Because the selected population showed an initial increase in zinc
tolerance and similiar laboratory studies have produced an increase in metal
tolerance (Blanc 1973, Burger 1973), it was concluded that fish possess  the
genetic potential to evolve metal-tolerant populations.  Laboratory studies
are usually limited to observing a few generations and are often subject to
confounding influences (e.g., inbreeding).  Hence, ultimate resolution of
whether fish can genetically adapt to increased metal levels will come from
studies of populations chronically exposed to metal contaminants.

     No studies have documented whether natural fish populations in
chronically contaminated waters ever realize this genetic potential.  Fish
are reported to live in a series of Canadian lakes receiving metal inputs
from nearby smelters (Van Loon and Beamish 1977), but a more recent study
indicates that long-term survival of these fish is uncertain.  McFarlane and
Franzin (1978) reported that a population of white suckers, Catostomus
commersoni, suffered severe reproductive impairment in one of  these lakes
with high zinc levels (141 to 341 mg/liter) although adult survival was
relatively unaffected.  Similar effects were noted on other fish species
from the same lake.  The smelter has operated since 1930, and  metal
concentrations presumably have been increasing since then, although
historical data on metal levels in the water are unavailable.  The onset of
harmful effects on local fish populations indicates that the fish failed to
adjust genetically to stressful metal levels during this relatively brief
period.

     A similar situation exists in the soft-water lakes of the Adirondacks,
where aluminum leached from surrounding soils by acid precipitation has
killed many fish (Cronan and Schofield 1979).  The development of harmful
metal concentrations in these lakes is a recent event, and again fish are
not adjusting rapidly to this sudden change in their environment.

     Mclntosh and Bishop (1976) used bluegills from a metal-contaminated and
an uncontaminated lake to compare relative survival in an acutely lethal
exposure of cadmium.  They found no difference in the 96-h LCcQ value for
the populations.  In a sublethal exposure of cadmium, they reported a
significantly lower cough rate for fish from the contaminated  lake than  for
control fish; no difference was found in breathing rates, however.  Whether
fish from the contaminated lake actually represented a population exposed to
selective pressures for metal tolerance is questionable, however, because
metals in the lake were not distributed evenly and it was not  known how  long
fish had been in the contaminated areas.

     This study focused on the potential for the rapid evolution of metal
tolerance in fish populations.  The results suggest that fish  possess this
potential, but the limited studies of natural fish populations living under
chronic metal stress do not support our findings.  Further work on fish
populations inhabiting waters with high trace-element levels is needed
before we can determine if and how fish can adapt ultimately to these
conditions.  Man's pollution of aquatic systems is extremely rapid on an


                                   -69-

-------
evolutionary time scale.  If this pollution continues, fish populations must
adapt at an equally rapid pace to avoid decimation or local extinctions.
                                    -70-

-------
                                  SECTION 5

          USE OF TEMPERATURE PREFERENCE AND ACTIVITY AS A SUBLETHAL
              BIOASSAY FOR THE  TOXIC EFFECTS OF ZINC TO BLUEGILL
INTRODUCTION

     The fly ash emitted from coal-fired generating  stations  contains
microcontaminant.s, including metal ions, that dissolve and may enter natural
water systems.  To monitor biological reactions  to these  toxicants, we  need
methods that are quick, sensitive, and accurate.  Such methods aid in
setting standards, monitoring spills, discovering synergistic effects  (the
cooperative effect of several factors working independently), and monitoring
microcontaminant levels in mining runoff and industrial waste waters.

     This section documents our effort to test one of these methods:   A
temperature-preference apparatus.  Preferred temperature  is a stable
behavioral trait for fish (Magnuson and Beitinger 1978).  Change in
temperature preference indicates a response to some  other factor such  as
stress from starvation (Javaid and Anderson 1967, Stuntz  1975, Stuntz  and
Magnuson 1976) or pollution (Ogilvie and Anderson 1965, Peterson 1973).  In
fact, knowledge of the concentration at which contaminants affect fish
temperature preference might serve as a useful indicator  of sublethal
toxicity.  Behavioral tests are a more sensitive indicator of these harmful
effects than lethality experiments and require less  time  and  space than
tests on sublethal chronic effects (Schere 1977, Henry and Atchison  1979).

DESIGN OF THE STUDY

     We tested an electronically controlled, temperature-selection apparatus
(Neill et al. 1972) for possible use as a tool for the detection of
sublethal concentrations of zinc by the bluegill (Lepomis macrochirus).  This
system uses a temporal gradient and provides a record of  activity and
temperature.  Zinc was chosen as the contaminant because  its  acute and
sublethal toxicity are well documented (Cairns and Scheier 1957, Sprague
1968, Brungs 1969, Burton et al. 1972, Waller and Cairns  1972, Cairns  et al.
1973).  A sublethal zinc concentration (2.5 ppm) was selected on the basis
of preliminary experiments and previous studies  (Burton et al. 1972, Waller
and Cairns 1972, Cairns et al. 1973).
                                    -71-

-------
MATERIALS AND METHODS

Selection Apparatus

     Nine aquaria, each divided into  two  compartments  with an
interconnecting tunnel, were equipped with a  system  that  permits  fish to
control the temperature of  the aquaria.   With a  constant  2°C temperature
difference between the two  compartments,  the  fish  controls the  direction of
temperature change (3°C/h)  in the  tank.   If  the  fish is  in the  warmer
compartment, the temperature of the whole tank is  increasing; if  the fish is
in the cooler compartment,  tank temperature  decreases.  Temperature is
selected oa a temporal rather than spatial basis  (Neill  and Magnuson 1974,
Beitinger et al. 1975).  A  computer continuously monitors and records tank
temperatares and movement through  the tunnel  (activity).   Changes due to
zinc may occur in the activity rate or selected  temperature (e.g., increase,
decrease, or diurnal pattern modification).

General Conditions

     The water used for the experiments was  Madison, Wis., city water
diluted at a ratio of 1:7 with distilled  water.   Madison  city water is
unusually hard (300 ppm CaCOo), and zinc  is  less toxic to the bluegill in
hard water  (Cairns and Scheier 1957). The 1:7 dilution results  in hardness
of 40 to 50 ppm (Table 18), a level within the range of  much previous v;ork
with zinc (Cairns et al. 1973).  Experiments  were  conducted during November
and December 1976.  The 12:12 light-dark  cycle included  periods of
intermediate light levels at dawn  and dusk.   We  analyzed  data from eight
control fish and 10 zinc-exposed fish.  Data were  not  utilized  if a given
fish did not pass through the tunnel  at least three  tlaes 'luring  a day or
night period.  Temperature  data from  one  fish were lost  because of
thermistor malfunction.

       TABLE 18.  ROUTINELY DETERMINED CHARACTERISTICS OF WATER USE IN
                     THE TEMPERATURE-PREFERENCE  BIOASSAY

Water
characteristic
pH3
Total hardnessb
Alkalinity0
Dissolved oxygen
Conductivity



Unit
__
ppm
ppm
% saturation
pmhos/cm at 25°C


No. of
analyses
250
114
114
68
250



Median
7.6
44
66
101
102



Range
7.1-8.0
24-58
30.90
96-105
73-121

aFisher pH meter  Model  150.
bEDTA Titrametric method  (American Public Health Association et al. 1975).
GMethyl Orange  indicator  nethod  (American Public Health  Association et al.
  1975).
dYSI Model 54A.
eYSI Model 33,  S-C-T meter.

                                     -72-

-------
Fish

     Young bluegill (7 to 10 cm) were captured with beach  seines  and  fyke
nets in Lakes Wingra and Mendota in Madison, Wis., and placed in  holding
tanks for at least 1 month before an experiment.  The fish were fed trout
pellets once daily and were acclimated to experimental light and  water
conditions.  Fish were not fed during the selection experiments.

Procedure
     The experiment was started by placing one fish  in each aquarium.   One
side of each tank was set at 21°C, the other at 23°C.  After  1 day  of  fixed
temperatures, each fish was allowed to thermoregulate for  2 days  before zinc
was added to half of the tanks, selected at random.  Enough zinc  sulfate (Zn
SO,) dissolved in distilled water was added to both  compartments  to achieve
a concentration of 2.5 ppm zinc in half of the aquaria,  selected  randomly.
Zinc was added only once at the start of the experiment.   Distilled water
was added to the control aquaria.  Water in both compartments of  each
aquarium was analyzed for zinc daily (Table 19).  Fish were allowed to
thermoregulate for 4 more days.  At the end of the experiment, the  bluegill
were weighed, measured, and frozen for analysis of zinc  concentrations in
the gills, liver, and muscle (Table 20).

                   TABLE 19.  ZINC CONCENTRATIONS (ppm)  OF
                   WATER IN TREATMENT AND CONTROL AQUARIA3
                   Zinc-exposed
                      (n-97)
                                           Control
                                             (n=86)
Day
Median
Quartiles
Median
Quartiles
4
5
6
7
8
1.49
1.24
1.13
1.16
1.00
1.33-1.55
1.13-1.31
1.02-1.28
0.99-1.36
0.91-1.21
0.03
0.04
0.04
0.03
0.03
0.02-0.11
0.02-0.06
0.01-0.13
0.06-0.15
0.00-0.11

 Analyses were done by atomic absorption photospectrometry.

RESULTS

     After the zinc was added, its concentration in the water  continuously
decreased probably because of absorption on particulate matter and uptake by
the fish (Table 19).  Neither selected temperatures (Figure  19) nor  rates of
activity (number of tunnel crossings/h) (Figure 20) of control and zinc-
exposed fish differed significantly (Mann-Whitney U Test, p  <  0.05,  Siegel
1956).  Preferred temperatures on the day before zinc was added and  of  the
day zinc was added were not significantly different in either  the control or
the zinc-exposed fish (Wilcoxon signed ranks test).  There was a trend  for
                                   -73-

-------
           30
        Ul
        cc
           28
        HI
        Q
        HI
           26
(2.5 ppm  Zn
   added)
             +4-
                          23.8
    1
          N-8
Control O	O
                                                                       Zn
                                                                                N=9
                                             4         5
                                                  DAY
Figure  19.  Median selected temperatures of  bluegill in control aquaria and in aquaria treated with
           zinc in a  7-day experiment.  Vertical bars represent the 25th to 75th percentiles.

-------
           cc
           040
           DC
           LU
        CDC/)
        IS
             30
             20
           HI
        LU
I-
LJ_
O
              10
                                                          Control O	O
                                                                    N=8
                                                                         Zn
                                                                               N=10
                               ( 2.5 ppm Zn added)

                               	I        TI
                                                 DAY
Figure  20.  Median number of tunnel passes  per hour by bluegill in control aquaria and in aquaria
           treated with zinc in a 7-day experiment.  Vertical bars represent the 25th to 75th
           percentiles.

-------
   TABLE 20.  ZINC TISSUE CONCENTRATIONS (ppm) AT THE END OF THE EXPERIMENT
         FOR RANDOMLY SELECTED FISH FROM TREATMENT AND CONTROL TANKS3
                    Zinc-treated
                       (n-4)
                            Control
                             (n=4)
               Median
Range
Median
Range
Gill
Liver
Muscle
90.4
107.5
28.6
75.3-107.2
90.7-129.8
23.0-33.8
86.8
91.2
30.8
76.9-89.81
83.6-125.2
26.7-34.9

aAnalyses were  done  by  neutron activation  and  by  each median n=4.

the  zinc-exposed fish to  prefer a  lower  temperature,  but  this  lasted for
only 1 day.   The rate of  activity  of  the control  fish,  however,  was
significantly lower  (P  <^  0.5) on the  day distilled  water  was added  than on
the  previous  day.  This followed a trend of  decreasing activity  rates
(Figure  20).  Activity  rates of the zinc-^exposed  fish were  not significantly
lower after the addition  of  zinc;  in  fact,  the median was greater  than that
of the previous day.

DISCUSSION

     Our selection system does not  detect  sublethal effects  of zinc at lower
concentrations  than other methods  tested in water of  similar quality.
Cairns et al. (1973) detected sublethal  zinc concentrations  of 2 to 3  ppm  by
continuously monitoring bluegill movement patterns perceived by light-beam
interruptions,  and by measuring bluegill ventilation  rates.   Sprague (1968)
found no change in selected temperature  of Atlantic salmon  (Salmo salar) in
a horizontal  temperature  gradient  after  24 h of exposure  to 0.16 ppm zinc.

     The effect of temperature on  zinc lethality varies with the type  of
lethality test  and species of fish, but  seems  greater at  higher
temperatures.   Survival time of rainbow  trout  exposed to  zinc  decreases at
higher temperature (Lloyd 1960).   Temperature  stress  induced by  increasing
the  temperature at a rate of 1.5°C every 10 min reduced survival time  at a
concentration of 32 ppm zinc (Burton  et  al.  1972).  Also, at 5.6 ppm zinc
deaths occurred in 96 h at 30°C but not  at 20°C.  Pickering  and  Henderson
(1966) found  no significant difference in  toxicity  to fish  at  15° and 25°C,
but  the trend was for higher toxicity at higher temperatures.  Cairns  and
Scheier (1957)  found 100% survival at 18° and  at 30°C for overlapping  zinc
concentrations.
     Apparently, a bluegill that behaviorally reduced its temperature while
exposed to lethal zinc concentrations would increase its probablity of
survival.  A nonsignificant trend toward lower temperatures in the presence
of zinc was observed in both Iteterson's (1976) and the current work, perhaps
indicating that further experiments could determine which zinc concentration
results in a change in temperature preference.

                                      -76-

-------
     In conclusion, our method does not appear to be any more suitable as a
sensitive indicator of sublethal effects of metal ions than other methods.
                                    -77-

-------
                                  REFERENCES
Adelman, I. R., L. L. Smith, Jr., and  G.  D.  Siesennop.   1976.   Effect of
    size of gold fish and fathead minnows on use  of  pentachlorophenol as a
    reference toxicant.  Water Res.  10:685-687.

Albaugh, D. W.  1972.  Insecticide tolerance of two  crayfish populations
    Proaombarus aotue in south central  Texas.  Bull.  Environ.  Contain.
    Toxicol. 8:334-338.

American Public Health Association,  American Public  Works  Association, and
    Water  Pollution Control Federation.   1975.  Standard methods  for the
    examination of water and wastewater.  13th Ed.   New York.  874 p.

Andren, A., M. Anderson, N. Loux, and  R.  Talbot.   1977. Aquatic  chemistry,
    p.  19-35.  In:  Documentation of environmental change  related to the
    Columbia Electric Generating Station.   Institute  for Environmental
    Studies, Univ. Wisconsin-Madison,  Rep.  92, Madison, Wis.

Andrews, C., and M. Anderson. 1980.  Impacts of coal-fired power  plants on
    local  ground-water systems.  U.S.  Environmental  Protection Agency,
    Cincinnati, Ohio.  In Press.

Antonovics, J. , A. D. Bradshaw, and  R.  G. Turner.   1971.   Heavy metal
    tolerance in fish.  Adv. Ecol. Res.  7:2-85.

Bedford, B. L.  1977.  Changes in wetland vegetation associated with leakage
    from the cooling lake of a coal-fired power plant.  M.S. Thesis.  Univ.
    Wisconsin-Madison, Madison, Wis.  29 p.

Beitinger, T. L., J. J. Magnuson, W. H.  Neil, and  W.  R. Shaffer.   1975.
    Behavioral thermoregulation and  activity patterns in Leopomis
    ayanellue.  Anim. Behav. 23:222-229.

Bengtsson, B. E.  1974.  Effects of  zinc on the mortality  and reproduction
    of  the minnow Phoxinus phoxinus  L.  Arch. Environ. Contam.   2:342-355.

Blanc,  J.  M.   1973.  Genetic aspects of resistance  to mercury poisoning in
    steelhead trout (Salmo gair'dner>i).   M.S. Thesis.   Oregon State Univ.,
    Corvallis, Oreg.  33 p.

Brown,  B.  E.   1976.  Observations on the tolerance  of the  isopod  Asellus
    meri.di.anu6 Rac. to copper and lead.   Water Res.  10:555-559.
                                    -78-

-------
 Brown,  G.  W.,  Jr.   1976.   Effects of polluting substances on enzymes of
     aquatic organisms.   J.  Fish.  Res. Board Can. 33:2018-2022.

 Brungs,  W. A.   1969.   Chronic toxicity of zinc to the fathead minnow,
     P-Lmephalee  prometas  Rafinesque.   Trans. Ara. Fish. Soc. 98:272-279.

 Bryan G. W.   1976.   Some  aspects  of  heavy metal tolerance in aquatic
     organisms.   In:   A.  P.  M. Lockwood (ed.)  Effects of pollutants on
     aquatic organisms.   Soc.  Exp. Biol.  Sem.,  Ser. 2.  Cambridge Univ.
     Press, Cambridge,  England.

 Burger,  C. V.   1973.   Genetic aspects of lead  toxicity in laboratory
     populations  of guppies  (Poecilia reticulata).  M.S.  Thesis, Oregon State
     Univ., Corvallis,  Oreg.   35 p.

 Burton,  D. T.,  E. L.  Morgan,  and  J.  Cairns, Jr.  1972.  Mortality curves of
     bluegills  (Lepomis macroehirus Rafinesque) simultaneously exposed to
     temperature  and  zinc  stress.   Trans. Am. Fish. Soc.  101:435-441.

 Cairns,  J., Jr., J.  S. Grossman,  K.  L. Dickenson,  and E. H.  Herricks.
     1971.  The  recovery of  damaged streams. ASB Bull. 18:79-106.

 Cairns,  J., Jr., and  A.  Scheier.   1957.   The effects  of  temperature and
     hardness of  water upon  the  toxicity  of zinc to the common bluegill
     (Lepomis maerochirus  Rafinesque). Notulae  Natural   Acad. Nat. Sci.
     (Philadelphia) 299:1-12.

 Cairns,  J., Jr., R. E. Sparks,  W. T.  Waller.  1973.   The relationship
     between continuous biological monitoring and water quality standards for
     chronic exposure,  p.  383-402.   In:   G.E.  Glass  (ed.)  Bioassay
     techniques and environmental  chemistry. Ann Arbor Science Publishers,
     Inc., Ann Arbor, Mich.

 Carpenter, K. E.  1924.   A  study  of  the  fauna  of rivers  polluted by lead
    mining in the Aberyswyth  district of Cardiganshire.   Ann.  Appl. Biol.
     12:1-23.

 Cherry,  D. S., and R. K.  Guthrie.  1977.   Toxic metals in  surface waters
    from coal ash.  Water Resour. Bull.  13:1227-1236.

 Cherry,  D. S., R. K. Guthrie, J.  H.  Rodgers, Jr.,  J.  C.  Cairns, Jr., and K.
    L. Dickson.  1976.  Response  of mosquito fish  (Gambusia  affinis) to ash
    effluent and thermal  stress.  Trans.  Am. Fish.  Soc.  105:686-694.

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.

Crow, J. F.  1957.  Genetics  of insect resistance  to  chemicals.   Annu.  Rev.
    Entomol.  2:227-246.
                                    -79-

-------
Crow, J. F., and M. Kiiaura.   1970.  An  introduction  to population genetics
    theory.  Harper and Row,  New York.   591 p.

Drummond, R. A., W. A. Spoor, and G.  F.  Olson.   1973.   Some short term
    indicators of sublethal effects of  copper on  brook trout Salvelinus
    fontinalis.  J. Fish. Res.  Board  Can.  30:698-701.

European Inland Fisheries Advisory Commission.   1974.   Water quality
    criteria for European freshwater  fish.  Report on  zinc and freshwater
    fish.  Rome, Italy.

Ferguson, D. E.  1967.  The ecological  consequences  of pesticide resistance
    in fishes.  Trans. North  Am. Wildl.  Conf. 32:10-107.

Ferguson, D. E., and  C. C. Gilbert.   1967.  Tolerances of three species of
    Anuran amphibians to five chlorinated hydrocarbon  insecticides.   J.
    Miss. Acad. Sci.  13:135.

Franklin, D. R., and  L. L. Smith, Jr.   1963.  Early  life  history of  the
    northern pike, Esox lucius  L., with  special  reference to the factors
    influencing the numerical strength  of year-classes.  Trans. Am.  Fish.
    Soc. 92:91-110.

Glass, G. E.   1978.   Ecological effects  of  coal-fired  steam electric
    generating stations.  Third National Conference  Interagency
    Energy/Environment Research and Development  Program.   Washington, B.C.

Gordon, R. L.   1978.  The hobbling of coal  policy and  regulatory
    uncertainties.  Science 200:153-158.

Gorham, E.   1976.  Acid precipitation and its influence upon aquatic
    ecosystems:  An overview.   Water  Air Soil Pollut.  6:457-481.

Harrell, D. M.  1978.  Effects  on crayfish  of exposure to coal-ash effluent
    and to chromiun contaminated food.   M.S.  Thesis,  Univ. Wisconsin-
    Madison, Tladison, Wis.  134 p.

Hasler, A. D.,  A. T.  Scholz,  and R. M.  Horrall.   1978.  Olfactory imprinting
    and homing  in salmon.  Am.  Sci. 66:347-355.

Helmke,  Phillip, Wayne Robarge, Paula Burger, Myles  Schoenfield, Robert
    Koons, and  Glenn  Hanson.  (Unpublished)   Impacts of coal combustion on
    trace elements in the environment.   U.S.  Environmental Protection
    Agency, Cincinnati, Ohio.

Henry, M. G.,  and G.  J. Atchison.   1979.   Behavorial changes in bluegill
    (Lepomie maor>ochii°U8) as  indicators  of  sublethal effects of metals.
    Environ.  Biol. Fish 4:37-42.
                                     -80-

-------
Hunt, B. P., and W. F. Carbine.  1951.  Food of young  pike Esox luoius L.
    and associated fishes on Peterson's Ditches, Houghton Lake,  Mich.
    Trans. Am. Fish. Soc. 80:67-83.

Ives, D. B., and C. D. Besadny.  1973.  Final environmental  impact  statement
    for Columbia Generating Station of the Wisconsin  Power and  Light
    Company.  Wisconsin Dep. Nat. Res., Bur. Environ.  Impact, Madison, Wis.

Javaid, M. Y., and J. M. Anderson.  1967.  Influence of  starvation  on
    selected temperature of some salmonids.  J. Fis-h.  Res. Board Can.
    24:1515-1519.

Johnson, F. H.   1956.  Northern pike year-class strength and spring water
    levels.  Trans. Am. Fish.  Soc. 86:285-293.

Jones, J. R. E.  1938.  The relative toxicity of salts of lead,  zinc and
    copper to the stickleback  (Gaeteroeteus aouleatus  L.) and the effect of
    calcium on the toxicity of lead and zinc salts.   J.  Exp. Biol.  15:394-
    407.

Jones, J. R. E.  1958.  A further  study of the  zinc  polluted river
    Ystwyth.  J. Anim. Ecol. 27:1-14.

Kelso, J. R. M., and F. J. Ward.   1977.   Unexploited perch  populations of
    West Blue Lake, Manitoba,  and  their interactions.   J. Fish. Res.  Board
    Can. 34:1655-1669.

Kincaid, H. L.   1976.  Inbreeding  in rainbow trout  (Salmo gairdneri)•   J.
    Fish. Res. Board Can. 33:2420-2426.

Leland, H. V., D.  J. Wilkes, and E. D.  Copenhaver.   1976.   Heavy metals and
    related trace  elements.  J. Water  Pollut.  Control  Fed.  48:1459-1486.

Lloyd, R.   1960.   The  toxicity of  ZnSO^ to rainbow  trout.   Ann. Appl.  Biol.
    48:84-94.

Magnuson, J.  J. , and T. L.  Beitinger.   1978.   Stability of  temperatures
    preferred in Centrarchid fishes and terrestrial  reptiles.  In:  E. S.
    Reese and F. J. Lighter (eds.) Adaptations  in the aquatic and
    terrestrial  envrionments.  John Wiley &  Sons, New York.

Magnuson, J.  J., A. M. Forbes, D.  M.  Harrell,  and J. D. Schwarzmeir.
     1980.  Response of stream  invertebrates  to  an ashpit effluent.  U.S.
    Environmental  Protection Agency,  Cincinnati,  Ohio.  In  press.

Maki, A. W., and H. E. Johnson.   1976.  Evaluation  of a toxicant on the
    metabolism of  model  stream communities.   J. Fish.  Res.  Board Can.
    33:2740-2746.

McCarraher, D. B.,  and R. E. Thomas.   1972.   Ecological significance of
    vegetation to  northern  pike, Esox  lucius,  spawning.  Trans. Am. Fish.
     Soc.  100:560-563.

                                    -81-

-------
McFarlane, G. A., and W. G. Franzin.   1978.   Elevated  heavy metals:   A
    stress on a population of white suckers,  Catostomus  aommereoni,  in
    Hamell Lake, Saskatchewan.  J. Fish.  Res.  Board  Can.   35:963-970.

Mclntosh, A., and W. Bishop.  1976.  Distribution and  effects  of  heavy
    metals in a contaminated lake.  Purdue  Univ., Water  Resour.  Res. Ctr. ,
    Tech. Rep. 85.  69 p.

McKim, J. M.  1977.  Evaluation of tests  with early  life stages  of fish for
    predicting long-term toxicity.  J.  Fish.  Res. Board  Can. 34:1148-1154.

McKim, J. M., R. L. Anderson, D. A. Benoit,  R.  L. Spehar,  and  G.  N.
    Stokes.  1976.  Effects of pollution  on  freshwater fish.  J.  Water
    Bollut.  Control Fed. 48:1544-1620.

McLean, R. 0., and A. K. Jones.  1975.  Studies  of tolerance to  heavy metals
    in the flora of the rivers Ystwyth  and  Clarach,  Wales.   Freshwater Biol.
    (G.B.) 5:431-435.

Neill, W. H., and J. J. Magnuson.  1974.  Distributional ecology and
    behavioral thermoregulation of fishes in  relation  to heated  effluent
    from a power plant at Lake Monona,  Wis.   Trans.  Am.  Fish.  Soc. 103:663-
    710.

Neill, W. H. , J. J. Magnuson, and G. D. Chipman.   1972.   Behavioral
    thermoregulation by fishes:  A new  experimental  approach.   Science
    176:1443-1445.

Ogilvie, D.  M., and J. M. Anderson.  1965.   Effect of  DDT on temperature
    selection by young Atlantic salmon, Salmo satar*  J. Fish. Res.  Board
    Can. 22:503-5'l2.

Peterson, R. H.  1973.  Temperature selection of Atlantic salmon (Salmo
    solar) and brooktrout (Salvelinus fontinalis) as influenced  by various
    chlorinated hydrocarbons.  J. Fish. Res.  Board  Can.  30:1091-1097.

Peterson, R. H.  1976.  Temperature selection of juvenile Atlantic salmon
    (Salmo salar) as influenced by various  toxic substances.  J. Fish. Res.
    Board Can. 33:1722-1730.

Pickering, Q. H., and C. Henderson.  1966.   The acute  toxicity of some heavy
    metals to different species of warm water fishes.   Int. J. Air Water
    Pollut.  10:453-463.

Priegel, G.  R., and D. C. Krohn.   1975.   Characteristics of a northern pike
    spawning population.  Wisconsin Dep.  Nat. Resour., Tech. Bull. 86.

Rachlin, J.  W., arid A.  Iterlmutter.   1968.  Response  of an inbred strain of
    platyfish and the minnow to zinc.   Prog.  Fish-Cult.  27:153-157.
                                    -82-

-------
Robertson, A.  1961.  Inbreeding in artificial selection  programmes.   Genet.
    Res. 2:189-194.

Scherer, E.  1977.  Behavior assays—principles, results,  and  problems.
    p. 33-40.  In:  Proc. 3rd Aquatic Toxicity Workshop,  Halifax,  N.S.
    Environmental  Protection Service, Tech. Rep. EPA-5AR-77-1.   Halifax,
    Can.

Siegel, S.   1956.  Nonparametric statistics for behavioral sciences.
    McGraw-Hill, New York.

Skidmore, J. F.  1964.   Toxicity of zinc  compounds  to  aquatic  animals with
    special  reference to fish.  Q. Rev. Biol.  39:227-248.

Smith, W. E.   1973.  A cyprinodonted fish.  Jordane'Lla flor*idae, as a
    laboratory animal for rapid chronic bioassays.   J.  Fish.  Res.  Board  Can.
    30:329-330.

Snedecor, G. W., and W.  G.  Cochran.  1967.  Statistical methods.  Iowa State
    Univ.   Press,  Ames,  Iowa.   593 p.

Sparks, R.  E., W.  T. Waller, and J. Cairns, Jr.   1972.   Effects of shelters
    on the  resistance of dominant and submissive bluegills Lepomis
    maerochirus  to a lethal concentration of  zinc.   J.  Fish.  Res.  Board
    Can.  29:1356-1358.

Spehar, R.L.   1976.  Cadmium and zinc toxicity to  flagfish Jordanella
    floridae.  J.  Fish.  Res. Board Can. 35:1939-1945.

Speranza, A. W., R. J.  Seeley,  V. A. Seeley,  and  A. Perlmutter.  1977.  The
    effect  of  sublethal  concentrations  of zinc on  reproduction in the zebra
    fish Brachydanion rerio Hamilton-Buchanan. Environ. Pollut. 12:217-222.

Sprague,  J.  B.   1968.  Avoidance reactions  of rainbow  trout to zinc sulphate
    solutions.   Water Res.  2:367-372.

Sprague,  J.  B.   1971.  Measurement of pollutant toxicity to fish.   III.
    Sublethal  effects and  "safe" concentrations.   Water Res.  5:245-266.

Stokes,  P.  M.,  T.  C. Hutchinson, and K. Krauter.   1973.  Heavy-metal
    tolerance  in algae  isolated from contaminated  lakes near Sudbury,
    Ontario.   Can.  J. Bot.  51:2155-2168.

Stuntz, W.  E.   1975.  Habitat  selection and growth of  bluegills.  Ph.D.
    Thesis,  Univ.  of Wisconsin-Madison, Madison,  Wis.   125 p.

Stuntz, W.  E.,  and J. J. Magnuson.   1976.  Daily  ration, temperature
    selection, and activity of  bluegill.   p.  180-184.   In:  G. W.  Esch and
    R.  W. McFarlane  (eds.)  Thermal ecology II. Energy Research and
    Development  Assoc.  Symp.  Ser. CONF-750425. Natl.  Tech. Int. Serv.,
    Springfield,  Va.
                                    -83-

-------
Swanson Environmental, Inc.  1977.  Cooling lake make-up water  intake
    monitoring program, March 1976-June  1977.  Wisconsin  Power  and  Light
    Co., Columbia Energy Center,  Portage, Wis., and  Southfield,  Mich.

U.S. Environmental Protection Agency.  1975.  Methods  for acute toxicity
    tests with fish, macroinvertebrates, and amphibians.  Ecol.  Res.  Ser. ,
    EPA 660/3-75-009, Corvallis,  Oreg.   61 p.

Van Loon, J. C., and R. J. Beamish.   1977.  Heavy-metal contamination by
    atmosphere fallout of several Flin Flon area lakes, and  the relation to
    fish populations.  J. Fish. Res.  Board Can. 34:899-906.

Vinson, S. B., C. E. Boyd, and D. E.  Furguson.  1963.  Resistance  to  DDT in
    the mosquito fish Gambueia affinis.  Science 139:217-218.

Waller, W. T., and J. Cairns, Jr.   1972.  The use of fish movement  patterns
    to monitor zinc in water.  Water  Res. 6:257-269.

Weatherly, A. H., J. R. Beevers,  and  P.  S. Lake.  1967.   The ecology  of a
    zinc polluted river.  In:  A. H.  Weatherly (ed.) Australian inland
    waters and their fauna.  Australian  National Univ. Press,  Canberra,
    Australia.

Webb, R. E., and F. Horsthall.  1967.  Endrin resistance  in  the pine
    mouse.  Science 206:1756-1762.

Williams, J. E.  1955.  Determination of age from the  scales of northern
    pike (Esox lucius L.).   Ph.D. Thesis.  Univ. of  Michigan,  Ann  Arbor,
    Mich.   185 p.
                                     -84-

-------
                                 APPENDIX A




          NUMBER OF FISH CAUGHT AT EACH SAMPLING STATION, 1976-78







      TABLE A-l.  NUMBER OF FISH CAUGHT AT EACH SAMPLING STATION,1976

Station
Species
Northern pike (Esox lucius)
Walleye (Stizostedion vitreum vitreum)
Muskellunge (Esox masquinongy)
Largemouth bass (Micropterus salmoides)
Rainbow trout (Salmo gairdneri)
Yellow perch (Perca flavescens)
Spotted sucker (Minytrema melanops)
Pirate perch (Aphredoderus sayanus)
White sucker (Catostomus commersoni)
Black crappie (Pomoxis nigromaculatus)
White crappie (Pomoxis annular is)
Rock bass (Ambloplites rupestris)
Pumpkinseed (Lepomis gibbosus)
Bluegill (Lepomis macrochirus)
Yellow bullhead ( Ictalurus natalis)
Black bullhead (Ictalurus melas)
Redhorse (Moxostoma sp.)
Bowfin (Amia calva)
Carp (Cyprinus carpio)
Buffalo (Ictiobus cyprinellus)

1
1
0
0
0
2
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
2
21
2
0
0
1
0
2
0
12
20
1
5
0
0
2
3
1
8
many
many
3
8
3
6
1
0
0
9
3
3
0
0
0
22
4
0
12
0
2
many
many
4
18
1
3
0
0
1
2+a
0
3
1
0
0
2
0
0
11
0
0
many
many
+ signifies more than two, but exact number not recorded.
                                   -85-

-------
TABLE A-2.  NUMBER OF FISH CAUGHT AT EACH SAMPLING STATION, 1977

Station
Species
Northern pike (Esox lucius)
Walleye (Stizostedion vitreum vitreum)
Muskellunge (Esox masquinongy)
Largemouth bass (Micropterus salmoides)
Rainbow trout (Salmo gairdneri)
Yellow perch (Perca flavescens)
Spotted sucker (Minytrema melanops)
Pirate perch (Aphredoderus sayanus)
White sucker (Catostomus commersoni)
Black crappie (Pomoxis nigromaculatus)
White crappie (Pomoxis annularis)
Rock bass (Ambloplites rupestris)
Pumpkinseed (Lepomis gibbosus)
Bluegill (Lepomis macrochirus)
Yellow bullhead (Ictalurus natalis)
Black bullhead (Ictalurus melas)
Redhorse (Moxostoma sp.)
Bowfin (Amia calva)
Carp (Cyprinus carpio)
Buffalo (Ictiobus cyprinellus)
White bass (Morone chrysops)
Golden shiner (Notemigonus crysoleucas)
Chestnut lamprey ( Ichthyomyzon castaneus)
Freshwater drum (Aplodinotus grunniens)
1
10
0
0
0
2
0
0
0
11
0
0
2
0
0
0
0
1
1
0
0
0
0
0
0
2
45
4
30
7
7
3
54
0
69
25
0
20
40
4
4
3
0
16
5
0
1
1
3
1
3
2
0
0
0
0
0
0
0
0
0
0
0
0
1
0
1
0
0
0
0
0
0
0
0
4
7
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
2
0
0
0
0
0
0

                             -86-

-------
TABLE A-3.  NUMBER OF FISH CAUGHT AT EACH SAMPLING STATION, 1978

Station
Species
Northern pike (Esox lucius)
Walleye (Stizostedion vitreum vitreum)
Muskellunge (Esox masquinongy)
Largemouth bass (Micropterus salmoides)
Rainbow trout (Salmo gairdneri)
Yellow perch (Barca flavescens)
Spotted sucker (Minytrema melanops)
White sucker (Catostomus commersoni)
Pirate perch (Aphredoderus sayanus)
Black crappie (Pomoxis nigromaculatus)
White crappie (Pomoxis annularis)
Rock bass (Ambloplites rupestris)
Pumpkinseed (Lepomis gibbosus)
Bluegill (Lepomis macrochirus)
Yellow bullhead (Ictalurus natalis)
Black bullhead (Ictalurus melas)
Redhorse (Moxostoma sp.)
Bowfin (Amia calva)
Carp (Cyprinus carpio)
Buffalo (Ictiobus cyprinellus)
Brown trout (Salmo trutta)
White bass (Morone chrysops)
Freshwater drum (Aplodinotus grunniens)
1
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
2
101
1
0
0
2
0
42
40
0
20
0
2
0
3
11
2
0
14
21
0
1
1
6
3
7
0
0
0
0
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
4
12
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
                              -87-

-------
                                  APPENDIX B

                 MARSHES  NEAR THE COLUMBIA GENERATING STATION


           TABLE B-l.   MARSHES NEAR THE COLUMBIA GENERATING STATION


Marsh area                                    Description
Station site
(after construction)
Duck Creek
Rocky Run A
Rocky Run B
Corning—Weeting Lakes
The area north of County J, west of the  Chicago,
Milwaukee, St. Paul and Pacific Railroad tracks,
south of Duck Creek.  Includes the mouths  of
Rocky Run and Duck Creek.

Includes all wetlands along Duck Creek east of
the Chicago, Milwaukee, St. Paul and  Pacific
Railroad tracks, up to the dam located along  State
Hwy 22-24 at Wyocena, Wis.

Wetlands south of County Hwy J, east  of  the
Chicago, Milwaukee, St. Paul and Pacific Railroad
tracks and west of State Hwy 51.   Includes
wetlands drained by the mint drain and by
Rocky Run Creek.  Fyke nets set in both  the mint
drain and Rocky Run showed few northern  pike  moved
this far upstream to spawn.

Wetlands along Rocky Run Creek east of State  Hwy 51
including Mud Lake.  Unlikely to be pike spawning
habitat for reasons given for Rocky Run  A.

Wetlands associated with Corning and  Weeting  Lakes
located north of  the Wisconsin River and
west of Portage, Wis.  Accessible  by  a small
creek flowing 8 km (about 5 miles) south of the
Wisconsin River.  Ground survey in August 1978
showed no obstructions to fish movement, but  very
shallow stream flow (4 to 5 cm deep,  1 m across) in
upper reaches of the creek.

                                    (continued)
                                     -88-

-------
TABLE B-l (continued)

  Marsh area
                    Description
Powers Creek
Whelen Bay
Hinkson-Rowan Creek
Lodi Marsh
O'Kee Bay
South Dekorra
Inlet
Merrimac Inlet
Prentice Creek
Baraboo River Mouth
The mouth of Powers (Rowen) Creek east to
Interstate 90-94.  Includes a portion of
Lake Wisconsin known as Whelen Bay.

Upstream tributaries of Powers Creek starting
at Interstate 90-94 eastward to stream headwaters.

Includes wetlands along Spring Creek from
its mouth at the Wisconsin River to marsh
upstream of the town of Lodi, Wis.
Ground survey in August 1978 indicated two small
spillways of about 0.5 m in the town of
Lodi which would prevent upstream migration except
during spring floods.  Fish would have to migrate
about 8 km (5 miles) upstream through the town
of Lodi to reach suitable spawning habitat.

Wetlands associated with a bay of Lake Wisconsin
east of O'Kee, Wis., and a bay
east of Pine Bluff at Harmony Grove, Wis.

Wetlands associated with a small stream south of
Dekorra, Wis., and just south of where Interstate
90-94 crosses the Wisconsin River.

A bay of Lake Wisconsin and associated wetlands
south of Merrimac, Wis.  State Hwy 113-78 and
the Chicago-Northwestern Railroad tracks
cross the bay, but do not prevent access
by spawning fish.

Wetlands associated with Prentice Creek which
joins the Wisconsin River north of Merrimac, Wis.

Wetlands located at the mouth of the Baraboo
River upstream to where Interstate 90-94
crosses the river.
                                    -89-

-------
                                  APPENDIX C

   REVIEW OF LITERATURE ON ENTRAINMENT FROM COOLING LAKE INTAKE STRUCTURES
     In the appendix, the possible entrainment damage to fish and
invertebrate populations at the Columbia site is discussed; the possible
damage appears minimal.  In addition, a 1978 study of fish entrainraent  at
the site by Swanson Environmental, Inc., has revealed that fish loss  due to
the present water-intake systems is minor.

     The effects of cooling-water intake on aquatic systems have been
studied at many power plants over the last 20 years. Although the  studies
differed in their approach, detail, and conclusions, four general  areas of
concern have emerged:   (1) Removal of animals suspended or swimming in  the
water column; (2) mechanical injury because of impingement upon intake
screens or abrasion in  pumps, pipes, and condensers; (3) the toxic effects
of biocides used in reducing the fouling of pipe systems by microorganisms;
and (4) the various effects of thermal shock during condenser passage.

     The removal of animals from the water column, including the impingement
of adult and juvenile fish, has become the focus of a federally mandated
monitoring program, pursuant to the requirements of Public Law 92-500.
Freeman and Sharma (1977) have conducted a survey of these programs,  but a
summary volume is not complete.  The removal aspect of cooling-water  intake
is relevant to the Columbia site; mechanical, toxic, and thermal aspects of
entrainment do not apply.  The Columbia station withdraws water from  the
artificial cooling lake to cool the superheated steam in the turbines.   It
is essentially a closed system, except that evaporative losses from the lake
require a constant input from the Wisconsin River.  The "make-up"  water is
presently pumped from the intake channel to the artificial lake by two
10,000-gal/min pumps.   Water is drawn down an intake channel that  connects
with the river approximately 3,000 ft from the cooling lake.  The  channel is
protected by two bar-grilles and a fish conservation traveling screen.

     Studies of mechanical injury and mortality during entrainment have been
reported by Marcy  (1973,  1976), Carpenter et al.   (1974),  Ginn et  al.
(1974), King (1974), Davies and Jensen  (1975), and  Polgar  (1975).   Several
reviews such as  those of Coutant  (1970) and Hillegas  (1977) have been
published.  Although survival of damaged organisms  is often quite  low,  it
does not appear  that the numbers of organisms lost  results  in serious
effects on the aquatic  systems.

     Biocides such as chlorine are usually used at  such low concentrations
that they pose no  threat to entrained organisms or  to the  receiving body of
water  (Marcy 1971, Bass and Heath  1975, Basch and  Truchan  1976,  Brungs  1976,

                                    -90-

-------
Seegert and Brooks  1977).  However, thermal  shock,  combined  with  small
amounts of chlorine, has a greater effect than either  increased temperatures
or chlorine levels  alone (Eiler and Delfino  1974, Ginn et  al.  1974).

     Cooling-systems designers now use predictive tools  to minimize
impact.  Curves and models predict for a given intake  design the  amount  of
mechanical damage (Polgar 1975) and the extent of lethal and sublethal
thermal effects (Coutant 1970) expected.  Models have  also been developed by
Goodyear (1977), Christenson et al. (1977),  and others to  forecast effects
of removal on given fish populations.

POTENTIAL EFFECTS OF COOLING WATER INTAKE AT THE COLUMBIA  SITE

     Effects of entrainment of aquatic organisms from  the  Wisconsin  River by
the Columbia Generating Station are different from  effects seen at most
other generating stations.  At Columbia there is no direct return of  the
entrained water to  the river.  The analogy of the intake acting as a  large
predator on the river ecosystem (Coutant 1970) is nore applicable than  in
"once through" cooling situations.  In assessing potential effects
researchers often draw a relationship between the percentage of water in the
river used and the  resulting effect on the river.   However,  organisms in
riverine communities typically show "patchy" distributions (Whitton  1975),
and larger organisms can either avoid the intake channel or  electively  swim
into it.

Zooplankton and Drifting Macroinvertebrates

     Zooplankton are too small to be screened out of the intake pumps and
are less able to avoid the influence of the  pumping current  than  are  larger
animals.  The percentage of total river flow removed by the  intake water at
Columbia presently  averages 0.3% with a maximum of  1.08%.  Assuming  that the
number of organisms entrained by the Columbia intake is proportional  to  the
volume of river water used, we expect no significant loss  of invertebrates
from the Wisconsin  River.  Several other entrainment studies at U.S.  power
plants (King 1974,  Davies and Jensen 1975, Hillegas 1977)  did  not
demonstrate measurable effects in downstream plankton  communities even where
abundant data were  available and generating  stations in question  diverted up
to 30% of the river flow.

Adult and Juvenile  Fish
     A 1-yr study of fish entrainnent at the Columbia  site  (Swanson
Environmental, Inc. 1977) reported the number, species,  length, and
reproductive condition of fish impinged on the temporary screen box  unit  and
on the traveling screen unit currently in use.   Sampling was  conducted  for  a
24-h continuous period once a week.  An estimated  14%  of the  total intake
volume was sampled.  The catch numbers were extrapolated to estimate  total
annual impingement as 668 + 387 fish/yr (mean +90%  confidence limits).   The
number of adult and juvenile fish impinged at Columbia is low, and even  if
all impinged fish die, no effect on the river system should occur.
                                    -91-

-------
Fish Eggs and Larvae

     The Swanson Environmental, Inc., study  (1977) also included  sampling
for fish eggs and larvae.  Submersible pumps were mounted behind  the
traveling screen unit and pumped the sample  water into 423-ym nets.   Pump
rates were sufficient to prevent fish from avoiding the sampler.   Estimated
annual entrainment of larval fish was 126,659 + 93,994 larvae/yr  (mean  +90%
confidence limits).  No northern pike or walleye larvae were caught in  the
samples.  According to a summary of fish-census data for the Columbia site
(Wisconsin Department of Natural Resources 1973), northern pike and walleye
spawn in the wetland adjacent to Duck Creek.  The mouth of Duck Creek is
located just upstream from the Columbia intake (Figure 1). Northern pike
larvae and fry remain on the spawning marshes until they attain a size  of
20 mm at 16-24 days after hatching (Franklin and Smith 1963).  Although
emigrating larvae of this size would not be  able to avoid the intake
current, the river currents may be strong enough in early spring  to sweep
larvae past the intake.  Larval walleye are  known to migrate from their
spawning marshes in intermittent pulses over a 10- to 15-day period (Priegel
1970).  By sampling once every 7 days, the period of walleye larval
entrainment could have been missed.  Walleye larvae may also avoid
entrainment by staying in the main currents  as they enter the Wisconsin
River, therefore bypassing the shoreline by  the intake.  Newly hatched
walleye larvae emerging from similar spawning situations on the Wolf  and  Fox
Rivers in Wisconsin tended to stay in the strongest currents until they
reached more lacustrine situations where zooplankton were abundant (Priegel
1970).

     In summary, as long as the Columbia intake continues to remove a small
percentage of the river flow, we expect no measurable effects of  entrainment
on the river system.  An exception might occur when organism distribution  is
patchy near the intake, and a significant portion of one year-class (e.g.,
walleye larvae) is entrained.  Aside from acting as a predator by removing
organisms from the Wisconsin River, the usual types of entrainment effects
(mechanical, toxic, and thermal) do not apply to the Columbia station.

BIBLIOGRAPHY FOR ENTRAINMENT

Basch, R. E., and J. G. Truchan.   1976.  Toxicity of chlorinated  condenser
    cooling waters to fish.  EPA-600/3-76-009, EPA Environmental  Research
    Laboratory, Duluth, Minnesota.

Bass, M. L., and A. G. Heath.  1975.  Toxicity of intermittent chlorine
    exposure to bluegill sunfish, Lepomis maorooh-irus'   Interaction with
    temperature.  ASB Bull. 22:40.

Brungs, W. A.  1976.  Effects of wastewater  and cooling water chlorination
    on aquatic life.  EPA-600/3-76-098, EPA  Environmental Research
    Laboratory, Duluth, Minn.

Carpenter, E. J., B. B. Peck, and  S. J. Anderson.  1974.  Survival of
    copepods passing through a nuclear power station on northeastern  Long
    Island Sound, U.S.A. Marine Biol. 24:49-55.

                                    -92-

-------
Christenson,  S. W.,  D.  L.  DeAngelis,  and A. G.  Clark.  1977.  Development of
    a stock progeny  model  for  assessing power plant effects on fish
    populations.   In:   Webster van Winkle (ed.) Proceedings of the
    conference on  assessing  the effects of  power plant-induced mortality on
    fish populations,  Gatlinburg,  Tenn.  Pergamon Press,  Inc., New York.

Coutant, C. C.  1970.   Biological  aspects of thermal pollution. In:
    Entrainment and  discharge  canal effects.  Chemical Rubber Co. Grit. Rev.
    Environ.  Control 1(3):341-348.

Davies, R. M., and L.  D.  Jensen.   1975.  Zooplankton entrainment at three
    mid-Atlantic power  plants.   J.  Water Pollut. Control  Fed. 47:2130-2142.

Eiler,  H. 0., and  J. J.  Delfino.   1974.  Limnological and biological studies
    of  the effects of  two  modes of  open-cycle nuclear power station
    discharge on the Mississippi River (1969-1973).  Water Res. 8:995-1005.

Franklin, D.  R., and L.  L.  Smith,  Jr.   1963.  Early life  history of the
    northern  pike, Esox  lucius  L. ,  with special reference to the factors
    influencing the  numerical  strength of year-classes.   Trans. Am. Fish.
    Soc.  92:92-110.

Freeman, R. F., and  R.  K.  Sharma.   1977.   Survey of fish  impingement at
    power plants in  the  United  States.  Vol. II. Inland waters.  ANL/ES-56
    Vol. II.  Argonne  National  Laboratory,  Argonne, 111.   328 p.

Ginn, T. C., W. T. Waller, and  G.  L.  Lauer.  1974.   The effects of power
    plant condenser  cooling  water  entrainment on the amphipod Gammarus sp.
    Water Res. 8:937-945.

Goodyear, C.  P.  1977.   Assessing  the  impact of power plant mortality on the
    compensatory reserve of  fish  populations.  In:   Webster Van Winkle (ed.)
    Proceedings of the  conference  on assessing  the  effects of power plant-
    induced mortality on fish  populations.   Gatlinburg, Tenn.  Pergamon
    Press, Inc., New York.

Hillegas, J. M., Jr.   1977.  Phytoplankton  and  zooplankton entrainment.  A
    summary of studies at  power plants in the United States.  Paper
    presented at Savannah  River Ecological  Laboratory Symposium, Augusta,
    Ga. (Preliminary draft).

King,  J. R.   1974.    A study  of  power plant  entrainment effects on the
    drifting macroninvertebrates of the Wabash  River,  M.S. Thesis, De Pauw
    Univ., Greencastle,  Ind.

Marcy, B. C.  1971.  Survival of young fish in  the  discharge canal of a
    nuclear power plant.   J. Fish.  Res.  Board Can.  28:1057-1060.

Marcy, B. C.  1973.  Vulnerability  and survival of  young  Connecticut River
    fish entrained at a nuclear power  plant.   J.  Fish. Res.  Board Can.
    30:1195-1203.
                                    -93-

-------
Marcy, B. C.   1976.  Planktonic fish eggs and larvae  of  the  lower
    Connecticut River and the effects of the Connecticut  Yankee  Plant  In:
    D. Merriman and L. Thorpe (eds.) The impact of  a  nuclear power  plant.
    Connecticut River Ecological Study, Monogr. 1.  Am.  Fish.  Soc.  Bethesda,
    Md.

Polgar, T. T.  1975.  Assessment of near field manifestations  of power
    plants.  Induced effects on zooplankton.  In:   Proceedings of the  2nd
    Thermal Ecology Symposium, Augusta, Ga.

Priegel, G. R.  1970.  Reproduction and early life  history of  the walleye  in
    the Lake Winnebago region.  Wisconsin Dep. Nat. Resour.  Tech. Bull.  45.

Seegert, G. L., and A. S. Brooks.  1977.  The effect  of  intermittent
    chlorination on fish:  Observations 3 1/2 years,  17  species, and  15,000
    fish later.  Paper presented at the 39th Midwest  Fish and  Wildlife
    Conference, Madison, Wis.

Swanson Environmental, Inc.  1977.  Cooling lake make-up water intake
    monitoring program, March 1976 - June 1977.  Wisconsin Power and  Light
    Co., Columbia Energy Center, Portage, Wis., and Southfield,  Mich.

Whitton, B. A.  1975.  River ecology.  Studies in ecology, Vol.  2.  Univ.
    California Press, Berkeley and Los Angeles.

Wisconsin Department of Natural Resources.  1973.   Final  draft environmental
    impact statement for the Columbia Generating Station of  the  Wisconsin
    Power and  Light Company, Madison, Wisconsin.
                                     -94-

-------
                                  APPENDIX D

                  REVIEW OF  LITERATURE ON ACID PRECIPITATION
     Acid rainfall, the topic of this appendix, is not considered  a
potential problem for aquatic ecosystems at Columbia because of  the high
hydrogen-ion buffering capacity resulting from the calcareous nature  of the
drainage basin.

     Recent studies in both North America and Europe have documented  the
occurrence of acid rains with a pH ranging from 2.1 to 5.0  (Likens and
Bormann 1974; Beamish 1974, 1976; Dickson 1975; Schofield 1976).   Rainwater
is normally slightly acidic, with a pH of 5.7, as a result  of the
equilibrium reaction between atmospheric carbon dioxide and water  forming
carbonic acid (FUCO/).  Both natural and anthropogenic processes,  however,
can add three strong mineral acids, sulfuric, nitric, and hydrochloric, to
atmospheric water with a resulting sharp decrease in pH (Gorham  1976).  The
most predominant of these acids is sulfuric (I^SO^), which  can be  formed in
substantial amounts from the sulfur dioxide (862) produced  as sulfur  in
fossil fuels oxidizes during combustion.  Coal normally has between  1 and 3%
sulfur, but the percentage can go as high as 6%.  Of less importance  are
nitric acid (HNOo) and hydrochloric acid (HC1), which are also produced by
fossil-fuel combustion through the oxidation of organic nitrogen and
chlorine, respectively.  These acids may then enter aquatic systems  through
rainfall or, in northern latitudes, through spring ice and  snow  runoff.

     The work of Cogbill and Likens (1974) illustrates that acid
precipitation is likely to remain a problem in certain areas.  By  graphing
isolines of rainfall pH falling over the eastern U.S., they have shown a
dramatic increase in the geographic area affected by acid rain,  as well as
an increase in rainfall acidity for the 10-year period 1956-66.

     The initial effects of acid input into lakes and streams depend  largely
on edaphic characteristics that determine their buffering capacity.   All
waters so far affected by acid precipitation have been in areas  that  are
geologically highly resistant to chemical weathering and usually have a low
concentration of major ions, particularly bicarbonate (HCO^), resulting in a
specific conductance less than 50 Jmhos/cm (Wright and Gjessing  1976).  Acid
rainfall into such weakly buffered systems causes a loss of bicarbonate ion
and its replacement by sulfate; hence sulfate is the major  anion in
acidified soft water, whereas bicarbonate predominates in non-acidified soft
water.  Acidified lakes are frequently found to contain elevated aluminum
and manganese concentrations that are attributed to dissolution  from
surrounding soils.  Elevated levels of other heavy metals (Pb, Zn, Cu, Ni)
                                    -95-

-------
may also exist downwind of major base-metal smelters  (Van  Loon  and  Beamish
1977).

     Ecological studies concerned with acidification  of aquatic ecosystems
have focused on fish poulations, since the loss of an exploitable fish
population is the most noticeable and economically important  consequence  of
acid precipitation.  Fish loss is reported to be a gradual process  resulting
not from acutely lethal pH changes, but rather from the failure to  recruit
new year-classes into the population (Beamish 1974).  At pH values  above  the
lower lethal level, interference with spawning has been demonstrated  in both
laboratory and field studies (Mount 1973, Beamish 1976).   The presumed
mechanism causing reproductive failure is disruption  of normal  calcium
metabolism that prevents females from releasing their ova  (Beamish  1976).
Long-term effects of acidification on fish populations were summarized by
Beamish (1975) as follows:  (1) failure to spawn, (2) low  serum Ca    levels
in mature females, (3) appearance of spinal deformities, (4)  decreases in
the average size of year-classes, (5) reduction in population size, and (6)
disappearance of species from lakes.

     Studies have indicated a genetic basis for acid  tolerance  at the
species level (Gjedrem 1976, Robinson et al.  1976, Schofield  1976)  and
selective breeding of acid-tolerant fish strains has  been  proposed  as a
means of stocking waters that have lost their natural populations.  The
observed rates of population extinction indicate, however, that
acidification has been proceeding too rapidly for natural-selection
processes to be effective in maintaining fish populations  under natural
conditions.

     Equally as serious as damage to fish are the less conspicuous  effects
of acid rain on aquatic organisms such as microdecomposers, primary
producers, zooplankton, and zoobenthos.  Studies in six Swedish lakes, where
the pH decreased by 1.4 to 1.7 pH units in the last 40 years, have
demonstrated an inhibition of bacterial decomposition with a  resultant
abnormal accumulation of coarse organic detritus (Hendrey  et  al.  1976a).
Rooted macrophytes, zooplankton, and benthic  invertebrates are  also stressed
by acidification of waters (Hendrey et al. 1976b).  Some of the effects of
pH on aquatic organisms are summarized in Table D-l.

     Although pH measurements of rainfall in  the vicinity  of  the Columbia
Generating Station have not been made, it appears unlikely that acid
rainfall will noticeably affect nearby aquatic ecosystems  for the following
reasons:  (1) The Wisconsin River, Rocky Run  Creek, and nearby  waters are
well-buffered systems with total alkalinities in the  range of 80 to 133
mg/liter CaCO-j and conductivities of 178 to 273 ^mhos/cm;  (2) winds are
predominately from the west and south (Stearns et al. 1977) and,  therefore,
power plant emissions should miss most of the nearby  aquatic  systems  that
are located mainly west and south of the plant; (3) the present pH  values of
the Wisconsin River (7.6 to 8.2) and Rocky Run Creek  (7.6  to  8.2) are well
within the recommended safe range of 6.5 to 9.0 for natural waters  and have
not changed noticeably since the plant began  in 1975.
                                    -96-

-------
            TABLE  D-l.   SUMMARY OF pH EFFECTS ON AQUATIC ORGANISMS
 pH
Effect
Reference
< 3.5       Unlikely that fish can survive for more
             than a few hours
            A few invertebrates (midges, mosquito,
             caddisfly) have been found
            Few plants (only mosses and algae) have
             been found
                            EIFAC 1969
                            Lackey (1938)
                            Hendrey et al.
                            1976a
3.5-4.0     Lethal to salmonids and bluegills, limit
             of tolerance of pumpkinseed, perch, and
             pike, but reproduction is inhibited
            Cattail (Typha) is the only higher plant

4.0-4.5     Only a few fish species survive, including
             perch and pike
            Lethal to fathead minnows
            Flora is restricted
            Some caddisflies and dragonfiles are found,
             and midges are dominant

4.5-5.0     Salmonids may survive, but do not
             reproduce
            Benthic fauna are restricted; mayflies
             are reduced
            Fish populations are severely stressed;
             a viable fishery is nonexistent
            Snails are rare or absent
            The fish community is decimated with
             virtually no reproduction
            White suckers and brown bullheads fail
             to spawn, but perch do spawn

5.0-6.0     Rarely lethal to fish except some
             salmonids, but reproduction is reduced
            Larvae and fry of sensitive species may
             be killed
            Bacterial species diversity is decreased,
             benthic invertebrates are reasonably
             diverse, but sensitive taxa such as
             mayflies are absent and molluscs are rare
            Fathead minnow egg production and ability
             to hatch are reduced
            Smallmouth bass, walleye, and burbot stop
             reproducing
            Roe of roach (Rutelus rutelus) fail
             to hatch

                                  (continued)
                            U.S. EPA  1973
                            U.S. EPA 1973
                            U.S. EPA 1973
                            Hendrey et al.!976a
                            Beamish 1974,  1975

                            Beamish 1975
                             U.S. EPA 1973
                             flount 1973

                             Beamish 1976

                             tlilbrink and
                             Johansson 1975
                                    -97-

-------
TABLE D-l (continued)
pH                          Effect                         Reference
6.0-6.5     Unlikely to be harmful to fish unless         U.S.  EPA 1973
             free CO^ exceeds 100 ppm
            Good invertebrate fauna except for
             reproduction of Gammarus and Daphnia
            Aquatic plants and microorganisms
             relatively normal

6.5-9.0     Harmless to fish and most  invertebrates
             although 7.0 is near the lower limit for     U.S.  EPA 1973
             Gammarus reproduction
            Microorganisms and plants are normal
            Toxicity of other substances
             may be affected by pH shifts within
             this range.
     Future considerations should be given  to  the  effect  of  added sulfur
emissions when Columbia II begins operation and on the  contributions,  if
any, of the Columbia plant emissions to acid rainfall over distant waters
such as northern Wisconsin lakes, some of which are  poorly buffered and  more
subject to acidification.

BIBLIOGRAPHY FOR ACID RAIN

Beamish, R. J.  1974.  Loss of  fish populations from unexploited remote
    lakes in Ontario, Canada as a consequence  of atmospheric fallout of
    acid.  Water Res. 8:85-95.

Beamish, R. J.  1975.  Long-term acidification of  a lake  and resulting
    effects on fishes.  Ambio 4(2):98-102.

Beamish, R. J.  1976.  Acidification of lakes  in Canada by acid
    precipitation and the resulting effect  on  fishes.   Water Air Soil
    Pollut.  6:501-514.

Cogbill, C. V., and G. E. Likens.   1974.  Acid precipitation in the
    northeastern United States.  Water Resour. Res. 10:1133-1137.

Dickson, W.  1975.  The acidification of  Swedish lakes.  Institute of
    Freshwater Research,  Drottningholm, Sweden.  Rep. No. 54.  p. 8-20.
                                    -98-

-------
European Inland Fisheries Advisory Commission Working  Party on Water
    Quality.  1969.  Water quality criteria for European  freshwater  fish:
    Extreme pH values and inland fisheries.  Water  Res.  3:593-611.

Gjedrem, T.  1976.  Genetic variation in tolerance  of  brown trout  to acid
    water.  SNSF-project, Norway, FR5/76.   11 p.

Gorham, E.  1976.  Acid  precipitation and  its influence upon aquatic
    ecosystems:  An overview.  Water Air Soil Pollut.  6:457-481.

Hendrey, G. R., K. Baalsrud, T. S. Traaen, M. Laake, and  G. Raddum.
    1976a.  Acid precipitation:  Some hydrobiological  changes.  Ambio 5(5-
    6):224-227.

Hendrey, G. R., R. Borgstrom, and G. Raddum.  1976b.   Acid precipitation in
    Norway:  Effects on  benthic faunal  communities.  Paper presented at the
    39th Annual Meeting, Am. Soc. Limn, and Oceonography, Savannah,  Ga.

Lackey, J. B.  1938.  The flora and fauna  of surface waters polluted by acid
    mine drainage.  Public Health Rep.  53:1499-1507.

Likens, G. E., and F. H. Bormann.  1974.   Acid  rain:   A serious regional
    environmental problem.   Science 184:1176-1179.

Milbrink,  G., and N. Johansson.  1975.   Some effects of  acidification on roe
    of roach, Rutilus rutilus L., and perch, Peraa  fluviatilis L., with
    special reference to the Avad System in eastern Sweden.  Institute of
    Freshwater Research, Drottningholm,  Sweden.   Rep.  No. 54.

Mount, D.  I.  1973.  Chronic effect of  low pH on  fathead  minnow survival,
    growth and reproduction.  Water Res. 7:987.

Robinson,  G. D., W. A. Dunson, J. E. Wright, and  G. E. Mamolito.  1976.
    Differences  in low pH tolerance among  strains of  brook trout (Salvelinus
    fontinalis)  J. Fish  Biol. 8:5-17.

Schofield, C. L.   1976.  Acid precipitation:  Effects  on fish.  Ambio 5(5-
    6):228-230.

Stearns, C. R.,  B. Bowen, and L. Dzamba.   1977.   Meteorology, p. 171-183.
    In:  Documentation of environmental change  related to the Columbia
    Electric Generating  Station.   10th  Semi-Annual  Progress Report.
    Institute for Environmental  Studies,   Univ. Wisconsin-Madison, Madison,
    Wis.   Rep. 82.

Van Loon,  J. C., and R.  J.   Beamish.   1977.   Heavy  metal contamination by
    atmospheric  fallout  of several Flin Flon area lakes,  and the relation to
    fish populations.  J. Fish Res. Bd. Can.   34:899-906.
                                     -99-

-------
Wright, R. F., and E. T. Gjessing.   1976.  Acid precipitation:   Changes  in
    the chemical composition of lakes.  Ambio  5(5-6):219-223.

U.S. Environmental Protection Agency.   1973.   Acidity,  alkalinity,  and pH,
    p. 140-141.  In:  Water quality  criteria.  Ecol.  Res.  Ser.,  R3-73-033.
                                    -100-

-------
                                  APPENDIX E

           REVIEW OF LITERATURE ON ALTERNATIVE DISPOSAL OF  FLY ASH


     Increased national emphasis  on  the use  of coal  to meet energy
requirements may result in a doubling of coal-ash production from  1975
levels by the year 1995 (PEDCO-Environmental, Inc.  1976).   Annual  coal-ash
production is currently estimated to be 61.9 x 10  tons (Davis and Faber
1977) and may be 100 x 106 tons by 1985 (Harriger 1977).  About  20% of the
ash is used for commercial purposes  in cement, asphalt and  concrete,
fertilizer, fire control, road-bed stabilizer, soil  aeration, and  sanitary
landfill cover (PEDCO-Environmental  1976, Theis  1976a, Harriger  1977).
Research continues into additional uses for  coal ash such as water
reclamation, sewage-sludge conditioning, and supplementation of  soil  sewage
micronutrients (Theis 1976a, Furr et al. 1977).  Fly ash and lime  cause
precipitation of phosphorus from natural waters, and the ash seals the
nutrients in the sediment; however,  the side effects of such treatment may
be severe (Theis and De Pinto 1976).  Fly ash concentrations of  10  to  20
g/liter were toxic to Stone Lake, Mich., fish.   High pH, dissolved oxygen
depletion, heavy-metal release, and  physical clogging and crushing of
organisms are other effects that  have not been adequately investigated.  Fly
ash applied to soils can neutralize  acid soils and supply calcium  and trace
elements  (PEDCO-Environmental, Inc.  1976); however,  the high conductivities
of fly-ash-water solutions may result in injuriously high salt
concentrations for many sensitive crops (Olsen and Warren 1976). Theis
(1976a) suggests the extraction of the following quantities of rare metals
from ash:  53.2 kg As/day, 5.2 kg Pb/day, 5.0 kg Cu/day, 49 kg Zn/day, 12.3
kg Cr/day, 730 g Cd/day, and 18.9 g  Hg/day.

     Despite continuing research  the large excess of fly-ash production over
demand is likely to continue (Theis  1976a) and, coupled with an  average rate
of ash production of 0.5 kg/kWh (PEDCO-Environmental, Inc.  1976),  will
result in large amounts of ash to be disposed of in  an environmentally sound
manner.  The new source performance  standards (NSPS) applicable  to new power
plants prohibit discharges into natural waters from  ash-settling ponds
(Dvorak and Pentecost 1977).  To comply with these regulations,  ash from
Unit II of the Columbia Generating Station is currently being held in a
segregated portion of the ash basin while a  site for permanent land disposal
is sought and prepared.

     Many concerns remain regarding  the landfill disposal of coal  ash. In
addition to the continued threat of  surface  contamination due to
precipitation and overland runoff, ground-water  contamination and  landfill
erosion are significant concerns.  Although many of  the principles of
sanitary landfilling are applicable  if consideration is given to the

                                   -101-

-------
different nature of the contaminants, an expanded study of coal-ash
landfills is needed.  Information on the leaching and mobility of ash  trace
constituents is limited (Dvorak and Pentecost  1977), and because  of  the
newness of the disposal method, little is known of the long-term  effects  of
such disposal.  Such studies are needed for the creation of  standards  for
land disposal of toxic sutbstances, which is virtually unregulated at  the
federal level (Fields and Lindsey 1975).

     The most widespread concern about coal-ash landfilling  is the potential
for ground-water contamination by leachate produced when water percolates
through the landfill.  High salt concentrations in leachate  may be a
significant problem, especially if it reaches  ground-water supplies  that  are
already high in salt.  Increased pH due to ash leachate may  be a  localized
problem (Olsen and Warren 1976), but pH is more important because of its
effects on metal solubilities and adscription.  This potential for metal  and
other trace-element contamination has received the greatest  attention  and
concern.

     The ability of the soil to attenuate contaminants in the leachate is of
primary importance in preventing ground-water  contamination  by any kind of
landfill.  Waldrip (1975) found that inorganic and organic materials from
sanitary landfill leachate are adsorbed by the soil, and many undesirable
ions are replaced by desirable ones in an ion-exchange process.   He
concluded that most ground-water contamination is limited to_ the  immediate
vicinity of the landfill because of slow movement of the ground water. The
low velocity allows sufficient time for ion exchange, dilution, and
dispersion to occur.  The landfill contribution to ground-water supply is
significantly diminished within a few hundred  feet of the landfill.

     Griffin et al. (1976) studied the attenuation of metals and  other
leachate constituents run through laboratory sediment columns.  Clay was
relatively poor in reducing concentrations of  Cl~, Na , and  water-soluble
organic compounds, but K, NH,, Mg, Si, and Fe  were moderately reduced  in
concentration, probably by cation exchange with Ca in the soil.   Low
leachate concentrations were strongly attenuated by small amounts of clay
possibly because of precipitation of the metals upon formation of metal
hydroxides or carbonates (caused by high pH and high bicarbonate
concentration in the^leachate).  Low leachate  concentrations of Al,  Cu, Ni,
Cr, As, SO,, and PO, precluded interpretation  for those substances.  Suarez
(1974) describes the chemical reactions involving metals leached  from
sanitary landfills and discusses their relationship with Eh, pH,  and
dissolved oxygen.

     A comparison of fly-ash landfill investigations is necessary to
determine the applicability of these sanitary  landfill results to the
landfill designed expressly for fly ash.  Theis (1976b) and  Theis and  Marley
(1976) discuss the potential for ground-water  contamination  from  land
disposal of fly ash.  They determined the important characteristics  of ash
to be initial trace-metal concentration, acid-base characteristics,  fly-ash
concentration in the aquatic system, and the size-fraction distribution of
the ash.  A combination of field and laboratory studies demonstrated that
Cr, Cu, Hg, Pb, and Zn were either released from leachate in insignificant

                                    -102-

-------
amounts or were rapidly sorbed onto soil particles.   The metals  As,  Ni,  and
Se, however, occurred in ground water at higher concentrations and appeared
able to migrate a greater distance. Sorptive processes  could  explain the
metal leachate behavior in the initial desorption of  metals from the ash
into water and subsequent adsorption onto the  soil phase.

     The investigation of a landfill for fly ash from combustion of  eastern
coal (Harriger 1977, Harriger et al. 1977) is  the most  comprehensive study
to date.  The presence of clay-rich soil was determined to be the most
important factor affecting water quality.  Other factors include composition
and quality of the ash, duration of exposure to leaching, pH, oxidation
conditions, and surface and ground-water flow  patterns.  Clay soils  were
relatively impermeable and found to adsorb or  exchange  large  quantities  of
ions.  Ground-water wells away from the landfill were lower in
concentrations of many trace substances, attesting to the benefits of
leachate percolation through the soil.  Landfill wells  often  had
concentrations of As, Se, Fe, Mn, and SO- above the U.S. Public  Health
Service drinking-water recommendations.  Landfill wells also  exhibited
higher concentrations of Zn, Ca, Cr, Cu, Mg, and K than the off-site
wells.  The metals Ca, Cr, and Cu were fairly  low, however, because  of  low
concentrations in the ash itself, good attenuation by clay, and  the
prevailing pH conditions.

     Analysis of surface waters  (streams flowing across the landfill, runoff
from the landfill, and ponds formed from precipitation) indicated few
effects of the landfill once the water left the site.   A stream  enclosed by
pipe as it crossed the site appeared to receive some  ground water and
ash-leachate seepage downstream.  Concentrations of Fe, Mn, and  SO^  exceeded
drinking-water standards, but decreased rapidly downstream.   Levels  of  Ca,
Cd, Cu, Fe, Mg, Na, Se, Zn, and  SO* were higher and pH  was lower in  ponds on
the landfill (especially those with exposed ash deltas) than  in  control
ponds away from the site.  Even higher concentrations of metals  occurred in
the sediments of the landfill ponds, indicating that  the contaminants were
precipitating out of the water.  Metal concentrations were high  in runoff
water from the landfill, and concentrations higher than in ground water  for
Cr, Cu, and Zn were evidence of attenuation by clay and restricted metal
mobility in ground water.  Thus surface runoff must be  contained to  permit
these mechanisms to operate.

     The pH and oxidation states of materials  in the  landfill influence  the
effectiveness of the attenuation mechanisms.   The solubility  of  most metal
ions is increased at lower pH values (Harriger 1977), and thus in acidic
leachate metals are not removed as readily by  the attenuation processes.
Generally, high pH greatly decreases solubility, and  only Zn  and Cd  are
considered soluble in the pH range 7 to 8.5 (Theis 1976a).  Most Cr  is
released from ash into the leachate at pH 3, although some is released  at pH
6, 9, and 12 (Theis and Wirth 1977).  Iron and Mn precipitate at pH  greater
than 7.5 (Harriger 1977).  Fields and Lindsey  (1975)  conclude that low  pH
affects ion exchange and adsorption properties of soil.  Clays are more
effective in adsorbing most metals when the pH is high, although a low  pH is
best for adsorption of organics.  They state that it  is best  to  maintain
landfill soils at pH 7.0-8.0.  Frost and Griffin (1977) found, however,  that

                                    -103-

-------
As and Se adsorption by clays is decreased at high pH.  Oxidation  causes  the
formation of iron oxides and hydroxides; these precipitate from the leachate
and can adsorb other ions (Harriger 1977), thus increasing the purification
capacity of the soil.

     The relative amounts of lime and amorphous iron oxides  in the ash
determine the pH of the leachate.  Western coals have high amounts of lime
(Theis and Wirth 1977), which account for the basic nature of the  ash from
the Columbia station.  The greatest environmental concern with ion pH ashes
is the large amount of surface leachable Fe  (Theis and Wirth 1977).  Theis
(1976a) states that a greater amount of metal is likely to be released  from
ash into ground water than into surface water, because of the lower pH  and
high CO^ content of ground water and the consequently greater likelihood  of
ion exchange from ash into this water.

     Research continues into the principles  of site selection and  design  to
reduce as much as possible the threat of ground- and surface-water
contamination.  Little is known about the potential environmental  effects of
landfills in Wisconsin (Zaporozec 1974) and  there have been  few long-term
studies of solid waste disposal in the  United States.  Leachate production
occurs even in well-designed landfills, especially in humid  areas  such  as
Wisconsin (Fields and Lindsey 1975, Zaporozec 1974), but this production  can
be minimized or controlled with proper  site  selection and design.

     Many investigators suggest the use of liners, either impervious to
retain all leachate, or permeable ones  to supplement the ability of the soil
to attenuate pollutants (Fields and Lindsey  1975, Griffin et al.  1976,
PEDCO-Environmental, Inc. 1976, Dvorak  and Pentecost 1977).  Where clay in
native soils is insufficient, a clay liner can satisfactorily mitigate  the
contamination threat.  It has been suggested that ash landfills nay have  the
capacity to seal themselves against leachate loss. As soluble CaO  moves into
the soil and forms CaCO-j, the permeability of the soil may be significantly
reduced (Olson and Warren 1976).  Fly ash is often deliberately applied to
sanitary landfills because of its moisture-adsorbing characteristics (PEDCO-
Environmental, Inc. 1976).

     Other suggestions to reduce the potential of contamination include
vegetating the landfill to reduce erosion by wind or water.  Harriger  (1977)
found that erosion remained a problem when the ash was covered with bare
soil.  PEDCO-Environmental, Inc. (1976) suggests the use of  species tolerant
to high pH, boron, and salt.  Recommendations for sanitary landfills in
southern Indiana include:  use of upland sites to avoid runoff from upland
areas, sites with soils or intervening  materials with high exchange and
adsorption capacities, and sites where  the water table is much below the
bottom of. the waste; use of leachate lagoons to prevent surface-water
contamination, and avoidance of areas subject to flooding (Waldrip and  Ruhe
1974).  PEDCO-Environmental, Inc. (1976) presents a detailed discussion of
geological, chemical, and engineering aspects of landfill site selection  and
design.  A literature review by Heidman and  Brunner (1976) lists  references
concerning site locations, investigation, monitoring, and management for
sanitary landfills.  Much of the information in both reports is also
applicable to coal-ash landfills.

                                    -104-

-------
     Several states and agencies have criteria and  regulations  that  should
be considered in the construction of coal-ash landfills  in  Wisconsin.   The
California State Water Resources Control  Board (1975)  lists the following:
(1) Underlying geological formations with questionable permeability  must  be
permanently sealed, or ground-water conditions must  prevent hydrologic
continuity; (2) leachate and subsurface flow must be self-contained; (3)
sites must not be located over zones of active faulting;  (4)  limitations  are
applied if the area is in a 100-yr (or more frequent)  flood-frequency  class.
The U.S. Environmental Protection Agency  (1973)  recommends  the  following
criteria:  (1) Low population density; (2) low alternate  land-use  value;  (3)
low ground-water-contamination potential; (4) away  from  flood plains,
excessive slopes, and natural depressions; (5) soil with  high clay content;
(6) adequate distance from human and livestock water supplies;  (7) areas  of
low rainfall and high evaporation rates, where possible;  (8)  sufficient
elevation over the water table; (9) no hydrologic connection  with  ground  or
surface water; (10) use of encapsulation, liners, waste  detoxification, or
solidification/fixation where necessary;  (11) adequate monitoring.
Consideration of all these suggestions will significantly reduce,  if not
avoid entirely, the adverse effects that a fly-ash  landfill might  have on
environmental quality.

     It appears that the high pH of the ash expected from Columbia II  will
substantially reduce the pollution potential from a landfill.   The landfill
site must be chosen carefully, however, to avoid direct  connection with the
ground water.  A clay or other type of liner will probably  be beneficial,  if
not required, to avoid ground-water contamination.   Pipes to  collect and
recirculate leachate should be used if there is any likelihood  of  less than
complete metal attenuation by the time the leachate reaches the ground
water.

SUMMARY

1.  Fly ash may be used commercially for a variety  of purposes  but supply
    will probably continue to exceed demand (Theis  1976a, PEDCO-
    Environmental, Inc. 1976, Theis and De Pinto 1976, Harriger 1977).

2.  Although recent air and water pollution standards prohibit  the discharge
    of ash or its leachate into surface waters, considerable  concern has
    arisen over the potential adverse effects of the dry  disposal  of fly  ash
    in landfills.

3.  Metal and trace-element contamination of water, particularly ground
    water, is the most serious concern.  Soils vary widely  in their
    abilities to attenuate these pollutants.

4.  Clay soils have the greatest capacity for metal adsorption  and ion
    exchange (Griffin et al.  1976,  Theis 1976b, Theis and Marley 1976,
    Harriger 1977).

5.  Because of these mechanisms, and the dilution and dispersion in  slow-
    moving ground water,  most  ground^water contamination is limited  to the
    immediate vicinity of the landfill (Waldrip 1975, Harriger  1977).

                                   -105-

-------
6.  With proper precautions direct surface-water contamination is  usually
    minimal (Harriger 1977).  Appropriate precautions include containment  of
    surface runoff and avoidance of low sites and steep slopes.

7.  Better attenuation of metals is usually obtained when the leachate has a
    high pH.  This is caused by the reduced solubility of metals and
    improved properties of clay under these conditions (Fields and Lindsey
    1975, Theis 1976a, Harriger 1977).  Fortunately, coal burned at the
    Columbia Generating Station produces basic conditions in its ash.

8.  Where natural soils are not sufficient, clay or impervious liners  should
    be applied to the landfill (PEDCO-Environmental, Inc. 1976, Dvorak and
    Pentecost 1977).  Fly ash appears to have some capacity to form a  seal
    itself (Olson and Warren 1976).

9.  Other recommendations to reduce the potential environmental
    contamination include covering with soil; encouraging vegetation;
    containing leachate; adequate monitoring; and avoiding sites with  high
    ground water; flooding potential; active faulting; or low elevations.

BIBLIOGRAPHY FOR FLY ASH

California State Water Resources Control Board.  1975.  Disposal site  design
    and operation information.  Sacramento, Calif,  p. 19-21.

Davis, J. E., and J. H. Faber.  1977.  Annual report:  National Ash
    Association.  National Ash Assoc., Washington, D.C.

Dvorak, A. J., and E. D. Pentecost.   1977.  Assessment of the health and
    environmental effects of power generation in the Midwest.  Vol. II.
    Ecological effects.  Draft.  Argonne National Laboratory, Argonne,
    111.  169 p.  (Permission obtained.)

Fields, T., and A. W. Lindsey.  1975.  Landfill disposal of hazardous
    wastes:  A review of literature and known approaches.  U.S.
    Environmental Protection Agency,  EPA/530/SW-165.  Cincinnati,  Ohio.
    36 p.

Frost, R. R., and R. A. Griffin.   1977.  Effect of pH on adsorption of
    arsenic and selenium from landfill leachate by clay minerals.   J.  Soil
    Sci. Soc. Am. 41:53-57.

Furr, A. K., T. F. Parkinson, P. A. Hinrichs, D. R. Van Campen, C. A.  Bache,
    W. H. Gutenmann, L. E. St. John,  Jr.,  I. S. Pakkala, and D. J. Lisk.
    1977.  National survey of elements and radioactivity in fly ashes.
    Environ. Sci. Technol. 11:1194-1201.

Griffin, R. A., K. Cartwright, N.  F.  Shimi, J. D. Steele, R. R. Ruch,
    W. A. White, G. M. Hughes, and R. H. Gilkeson.  1976. Attenuation  of
    pollutants in municipal landfill  leachate by clay minerals.   Part  1:
                                    -106-

-------
    Column leaching and field verification.   Illinois  State  Geol.  Surv.
    Environ. Geol. Notes, No. 78.  34 p.

Harriger, T. L.   1977.  Impact on water quality  by  a coal  ash landfill in
    north central Chautaqua County, New York.  Ph.D. Thesis,  State
    University College, Fredonia, N. Y.   192  p.

Harriger, T. L., W. M. Benard, D. R. Corbin,  and D. A.  Watroba.  1977.
    Impact of a coal ash landfill on water quality  in  north  central
    Chautaqua County, New York.  Symposium on Energy and Environmental
    Stress in Aquatic Systems.  Savannah  River Ecology Laboratory.
    (Abstracts).

Heidman, J. A., and D. R. Brunner.  1976.  Solid wastes and  water  quality.
    J. Water Pollut. Control Assoc. 48:1299.

Olson, R. A., and G. Warren.  1976.  Aquatic  pollution potential of fly ash
    particles, p. 91-112.  In:  Toxic effects  on the biota from  coal and oil
    shale development.  Nat. Res. Ecol. Lab.,  Colorado State Univ.,
    Internal Proj. Rep. No. 7, Ft. Collins, Colo.

PEDCO-Environmental, Inc.  1976.  Residual waste best  management
    practices:  A water planner's guide to land  disposal.  U.S.
    Environmental Protection Agency, EPA/440/9-76/022,  Cincinnati, Ohio.

SuareE, D. L.  1974.  Heavy metals in waters  and soils  associated  with
    several Pennsylvania landfills.  Ph.D. Thesis,  Pennsylvania  State Univ.,
    University Park, Pa.  222 p.

Theis, T. L.  1976a.  Potential trace metal contamination  of water resources
    through disposal of fly ash.  Notre Dame  Univ., CONF-750530-3, South
    Bend, Ind.  21 p.

Theis, T. L.  1976b.  Contamination of ground water by heavy metals from the
    land disposal of fly ash.  Tech. Prog. Rep.   1  June 1976 to  31 August
    1976.  Prepared for U.S. Energy Research  and Development Administration,
    Notre Dame Univ., South Bend, Ind.  44 p.

Theis, T. L., and J. V. DePinto.  1976.   Studies on the reclamation of Stone
    Lake, Michigan.  U.S. Environmental Protection  Agency, Ecol. Res.  Ser.,
    EPA-600/3-76-106,  Cincinnati, Ohio.  84  p.

Theis, T. L., and J. J. Marley.  1976.  Contamination  of ground  water by
    heavy metals from the land disposal of fly ash.  Tech. Prog. Rep.
    1 June 1976 to 29 February 1976.  Prepared for  U.S. ERDA, Notre Dame
    Univ., South Bend, Ind.  21 p.

Theis, T. L., and J. L. Wirth.  1977.  Sorptive  behavior of  trace  metals on
    fly ash in aqueous systems.  Environ. Sci.   Technol. 11:1096-1100.

U.S. Environmental Protection Agency.  1973.   Acidity,  alkalinity, and pH,
    p. 140-141.  In:  Water quality criteria.  Ecol. Res.  Ser. R3-73-033.

                                   -107-

-------
Waldrip, D. B.  1975.  The effect of sanitary  landfills  on water  quality in
    southern Indiana.  Ph.D. Thesis, Indiana Univ.,  Bloomington,  Ind.
    160 p.

Waldrip, D. B., and R. V. Ruhe.   1974.   Solid  waste  disposal  by land burial
    in southern Indiana.  Water Resour.  Res. Center,  Tech.  Rep.  No.  45.
    Purdue Univ., West Lafayette, Ind.   110 p.

Zaporozec, A.  1974.  Hydrogeologic evaluation  of  solid  waste disposal in
    south central Wisconsin.  Wisconsin  Dep. of Nat.  Resour., Tech.   Bull.
    78.  31 p.
                                     -108-

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
  EPA-600/5-80-078
                              2.
                                                            3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
   Ecological Studies of  Fish Near a Coal-Fired
     Generating Station and Related Laboratory  Studies
  Wisconsin Power Plant  Impact  Study	
             5. REPORT DATE
               July 1980 issuing  date
             6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
   John J. Magnuson,  Frank J. Rahel, Michael  J.  Talbot,
   Anne M. Forbes, Patrica A. Medvick
                                                            8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
   Department of Limnology
   University of Wisconsin
   Madison, WI  53706
             10. PROGRAM ELEMENT NO.

                   1BA820
             11. CONTRACT/GRANT NO.

                   R803971
12. SPONSORING AGENCY NAME AND ADDRESS
  Environmental Research Laboratory
  Office of Research and Development
  U.S.  Environmental Protection Agency
  Duluth, Minnesota  55804
              13. TYPE OF REPORT AND PERIOD COVERED
             14. SPONSORING AGENCY CODE

                EPA/600/03
15. SUPPLEMENTARY NOTES
16. ABSTRACT
 Construction of a coal-fired electric generating station on wetlands  adjacent to the
 Wisconsin River has permanently altered about  one-half of the original  1,104-ha site.
 Change in the remaining wetlands continues as  a  result of waste heat  and ashpit efflu-
 ent produced by the station.   Leakage of warm  water from the 203-ha cooling lake is
 causing a shift in the wetlands from shallow to  deep-water marsh.  Coal-combustion
 byproducts enter the wetlands from the station's ashpit drain.  Since this area was
 known to have a diverse fish community and to  be a spawning ground for  Wisconsin River
 game fish, we studied the  effects of this habitat loss and degradation  on fish popula-
 tions.  In laboratory experiments we investigated the use of temperature preference
 and activity as a sublethal  bioassay.  In selection experiments we examined the
 potential of fish to evolve  metal-tolerant populations in chronically contaminated
 environments.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                           c.  COS AT I Field/Group
 Thermal pollution
 Fish
 Sublethal effects
 Ashpit  effluents
 Wisconsin power  plant
   study
 Fish habitats
      06/F
      07/B
      07/C
18. DISTRIBUTION STATEMENT

  Release to  Public
19. SECURITY CLASS (This Report)
      unclassified
21. NO. OF PAGES
      121
                                              20. SECURITY CLASS (Thispage)
                                                     unclassified
                                                                          22. PRICE
EPA Form 2220-1 (Rev. 4-77)
                      PREVIOUS EDITION IS OBSOLETE
                                                          •ft U.S. GOVERNMENT PRINTING OFFICE:  1980--657-165/0076
                                            109

-------