PB81-150997
Effects of Fluctuating, Sublethal
Applications of Heavy Metal Solutions
Upon the Gill Ventilation Response of
Bluegills (Leporais macrochirus)
Virginia Polytechnic Inst, and State Univ.
Blacksburg
Prepared for
Environmental Research Lab.-Duluth
Cincinnati, OH
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
NTIS

-------
                                            EPA-600/3-81-003
                                            January 1981

                                            PP81-15CQP7
  EFFECTS OF FLUCTUATING,  SUBLETHAL APPLICATIONS
OF HEAVY METAL SOLUTIONS UPON THE GILL VENTILATION
    RESPONSE OF BLUEGILLS (Lepomis macrochirus)
                         by

                 John Cairns,  Jr.
                 Kenneth W. Thompson
                 Albert C.  Hendricks
              Biology Department and
    University Center for  Environmental Studies
Virginia Polytechnic Institute and State University
           Blacksburg, Virginia   24061
                    R 80 52740 10
                   Project Officer

              William B.  Horning, II
          Newtown  Fish Toxicology Station
             Cincinnati,  Ohio   45244
                      0V
                 NATIONAL TECHNICAL
                 INFORMATION SERVICE
                    OS. DEPARTED! OF COBSEBCE
                        itio, va 22161

-------
                NOTICE    Vr

THIS DO C UMi'NT HAS B E'EN RE PROD U C ED
F;R.OM THE  BE;ST" COP? FURNISHED us BY
THE; sPON s;oRI NG  A;G E-N-G^, "; A tTTOuGH'"IT•'_';
is  RECOGNIZED 'THAT CERTAIN  PORTIONS
ARE  ILLEGIBLE, IT IS BEING RELEASED
IN THE INTEREST  O-F/MAKING"AVAILABLE
AS MUCH INFORMATION  AS. POSSIBLE.

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
    EPA-600/3-81-003
3. RECIPIENT'S ACCESSION NO.
            L
4. TITLE AND SUBTITLE
 Effects of Fluctuating,  Sublethal Applications of Heavy
 Metal Solutions upon  the Gill Ventilation Response of
 Bluegills  (Lepomis macrochirus)
5. REPORT DATE
    January 1981  Issuing Date.
8. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
 J.  Cairns, Jr., K. W. Thompson,  and A. C. Hendricks
                                                           8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
 Biology Department and University Center for
   Environmental Studies
 Virginia Polytechnica Ins. and State University
 Blacksburg,  Virginia  24061
                                                            10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
  R805274010
12. SPONSORING AGENCY NAME AND ADDRESS
U.So  Environmental Protection Agency
Newtown Fish Toxicology  Station
3411  Church Street
Cincinnati,,  Ohio  45244
                                                            13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE

  EPA-600/03         •'••
15. SUPPLEMENTARY NOTES
16. ABSTRACT
The ventilatory response  of the bluegill to fluctuating,  sublethal amounts of heavy-  «
metals was investigated.   Non-contact  submerged,  stainless steel electrodes were
used to detect the weak electrical potentials that are produced when fish ventilate  '•'
 their gills.  These signals were amplified using high-gain amplifiers which were
interfaced with a minicomputer.  These ventilatory data were accumulated continuously,
and both the ventilatory  rates and average signal  amplitudes were recorded on
electronic tape for later analysis.

The response to these  toxic solutions was an increase in  rate as well as a decrease
in signal amplitude.   The latter response was shown to be a real response of the fish
and not due to changes in the electrical properties of the water when toxicant  was
added.  Due to the extreme smoothing of the data,  the analysis of variance was  unable
 to detect any differences between the toxicant application patterns.  It was signi-  -
ficant to note, however,  that the fish were capable of reacting to each subsequent
pulse of toxicant in a similar manner.
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.IDENTIFIERS/OPEN ENDED TERMS  C. COSATI Field/Group
18. DISTRIBUTION STATEMENT
 RELEASE TO PUBLIC
                                              19. SECURITY CLASS (ThisReport)
                                                UNCLASSIFIED
                                                                         21. NO. OF.PAGES
                                              20. SECURITY CLASS (This page)
                                                UNCLASSIFIED
                                                                         22. PRICE
EPA Form 2220-1 (Rov. 4-77)
                      PREVIOUS EDITION IS OBSOLETE

-------
                           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
or commercial products constitute endorsement or recommendation
for use.                 .        ...
                               ii

-------
                            FOREWARD


     The Federal Water Pollution Control Amendments of 1972, the
1977 Clean Water Act,  and the Toxic Substances Act all require
control of the discharge of toxic substances.  A reasonable data
base exists on toxicity of known chemical compounds to aquatic
organisms.  However, information is limited concerning the toxicity
or potential hazard of complex wastes that contain numerous
chemicals in varying concentrations and combinations.  Because
of the increasing number and complexity of chemical compounds
being manufactured and the fact that aquatic organisms are not  .....
usually exposed to individual chemicals, but rather to mixtures
of several or more, it is obvious that we cannot control toxic
discharges on a chemical-by-chemical basis.  Thus, a rapid      "'•"••..
biological test method is needed by regulatory agencies and dis-
chargers to evaluate the potential hazards of complex and variable
wastes prior to discharge into receiving waters.  The biological
system must integrate the effects of these complex mixtures over-
time.

     This report provides information concerning the acute
toxicity of several known heavy metals as well as mixtures of these
toxicants.  The report also provides information regarding the
use of fish respiratory impairment as a measure of toxicant stress
when subjected to sublethal fluctuating concentrations of indi-
vidual heavy metals and mixtures of these toxicants.
                                William B. Horning
                                Chief
                                Newtown Fish Toxicology Station
                                Cincinnati, Ohio
                               iii

-------
                             PREFACE

     The original proposal, forwarded January 28, 1977, listed
John Cairns, Jr. as the principal investigator.  Kenneth L.
Dickson and W. H. van der Schalie were to be- staff members on
the project.  Van der Schalie left before the project was funded
and was replaced by Kenneth W. Thompson when the project was
funded.  Dickson's responsibilities were taken over by Albert C.
Hendricks in June, 1978.
                               IV

-------
                            ABSTRACT

     The ventilatory  response of the bluegill  to  fluctuating,
 sublethal  amounts  of  heavy metals was  investigated.   Non-contact,
 submerged,  stainless  steel electrodes  were used to  detect  the
 weak electrical  potentials that are produced when fish  ventilate
 their  gills.   These signals were amplified using  high-gain
 amplifiers which were interfaced with  a minicomputer.   These
 ventilatory data were accumulated continuously, and both the
 ventilatory rates  and average signal amplitudes were recorded  on
 electronic tape  for later analysis.

     Acute toxicity of each of the heavy metal solutions used
 was determined as  well as the various  mixtures used.  The  com.......
 pounds used and  the LCSOs determined (in parentheses) were ZnS04
 (3.2 mg/1), CuCl2  (1.0 mg/1), NiCl2 (21.2 mg/1),  and K2Cr04    *
 (132.9 mg/1).  The toxicity of zinc-copper and zinc-copper-nickel-
 chromium mixtures  was found to be additive under  the test  condi-
 tions.

     Use of conventional parametric procedures is inappropriate
 for analysis of  ventilatory data from  an individual fish as
 successive observations are not independent.   This  report  deter-
 mined  that such  analyses are appropriate when  comparing data
 between individuals.

     The ventilatory  behavior of bluegills (Lepomis macrochirus)
 receiving  sublethal exposures of heavy metal solutions  was com-
 pared  to unexposed controls.  The toxicant solutions were
 delivered  in short, intermittent 5-hour pulses, longer  20-hour
 pulses, and conventional 96-hour applications.   .

     The response  to  these toxic solutions was an increase in
 rate as well as  a  decrease in signal amplitude.   The latter
 response was shown to be a real response of the fish and not due
 to changes in the  electrical properties of the water when  toxicant
 was added.  Due  to the extreme smoothing of the data, the  analysis
 of variance was  unable to detect any differences  between the
 toxicant application  patterns.  It was significant  to note,
 however, that the  fish were capable of reacting to  each subse-
 quent  pulse of toxicant in a similar manner.

...  	'Both  ventilatory rate and ventilatory amplitude, especially
 when used  in conjunction, can be powerful tools for the continuous
 monitoring of surface water quality.

-------
     This report was submitted in fulfillment of Contract No.
R 80 52 740 10 by the University Center for Environmental Studies
under the sponsorship of the U. S. Environmental Protection
Agency.  This report covers the period July 11,  1977 to April  11,
1980.
                               vi

-------
                           CONTENTS

Foreward	   iii
Abstract 	     v
Figures	viii
Tables	    xi
Acknowledgment 	   xii

   1.  Introduction  	     1
   2.  Conclusions 	     5
   3.  Materials and Methods 	     6
            Test specimens	     6
            Toxicity tests 	     6
            Water characteristics  	     9
            Ventilatory studies  	    12
   4.  Results and Discussion	    32
            Toxicity tests 	    32
            Ventilatory studies  	    34

References	    44
Appendices
   A.  Results of Water Analyses and Individual Toxicity
       Tests	    51
   B.  Graphic representation of the average Ventilatory
       rates for all 12 Ventilatory tests	    56
   C.  Graphic representation, of the average signal
       amplitude for all 12 ventilatory tests  ......    69
   D.  Examples of stressed and unstressed inter- and
       intra-individual variation as illustrated by strip
  •     chart recordings	    82
                               vii

-------
                             FIGURES

Number

   1      A bluegill in a ventilatory sensing chamber.
          Water flows into the receiving chamber (A),
          through the sensing chamber (B), and out the
          drain (C).  The electrodes are the S-shaped
          wires at either end of the sensing chamber.
          The signal is carried to an amplifier (AM) .

   2      An example of the amplified ventilatory
          signal of a bluegill ventilating at
          approximatly 20 times per minute ......
          The solenoid controlled, continuous flow,
          proportional diluter.  Th'e:' time delay! relay
          simultaneously opens ,the.v:valves on the head
          boxes and closes those on.-the proportioning
          boxes allowing the proportioning boxes to
          fill with diluent and toxicant solutions.
          When the level of the liquid in the diluent
          proportioning box reaches the top in
          compartment 5, a float valve is tripped and
          the solenoid valves are returned to their
          normal state,, delivering'toxicant and
          diluent first to the mixing chambers and
          ultimately to the fish chambers.  This also
          resets the time delay relay, and the cycle.
          is repeated  . . . . . .  . . . ...... ... .
          An individual "ventilatory sensing chamber.
          The dimensions of the fish holding chamber
          are: length, 12 icm; width,-4 cm; height,
          10 cm; and\water depth, 6 cm	13

          Cutaway view of the plywood isolation
          chamber to show the composition ventilatory
          sensing chambers	13

          The water delivery and data accumulation
          system	15

          The rate of increase of the concentration
          in the ventilatory sensing chambers   	  16


                              viii

-------
Number
   8      The rate of decrease of the concentration
          in the ventilatory sensing chambers  	  16

   9      Strip chart recording of the ventilatory
          signal of a bluegill.  A.  The ventilatory
          signal as taken directly from the amplifier.
          B.  The computer produced derivitive of
          the signal in A.  The very sharp negative
          peaks indicate that a count has been
          recorded	19

  10      Strip chart recording of the venilatory
          signal of a bluegill.  A.  The ventilatory
          signal as taken directly from the amplifier.
          Note the increased bimodality with respect
          to Figure 9.  B.  The computer produced
          derivitive of the signal in A	20

  11      The average ventilatory rate of a bluegill.
          The arrows indicate points at which the
          DC voltage drift caused a loss of signal.
          Replacing the electrodes and terminal
          barrier strips eliminated this problem   	  22

  12      Toxicant pattern 1.  Toxicant was delivered
          to fish 7 and fish 14 for 96 hr.  The
          shaded area on the fourth vertical plane
          indicates the shape,, duration, and concentra-
          tion of the application of the application
          of the toxicant.  Fish 9 was a control and
          received no toxicant.  Interval 2 = 0 - 96 hr
          (X axis); Interval 3 = 96 - 192 hr; Interval
          4 = 192 - 216 hr.  Data were not recorded
          during Interval 1 which was the 24 hr prior
          to hour 0 (See text for explanation	25

   13     Toxicant pattern 2.  Toxicant was delivered
          to fish 14 and fish 13.  The shaded area on
          the fourth vertical plane indicates the shape,
          duration, and concentration of the toxicant
          applications.  Each application lasted for
          20 hr; each increment was 4 hr.  Interval 2 =
          0 - 96 hr (X axis); Interval 3 = 96 - 156 hr;
          Interval 4 = 156 - 216 hr.  Data were not
          recorded during Interval 1 which was the 24 hr
          prior to hour 0 (see text for explanation)  ....  26
                               IX

-------
Number                                                      Page

  14      Toxicant pattern 3.  Toxicant was delivered
          to fish 16 and fish 8.  The shaded area on
          the fourth vertical plane indicates the
          shape, duration, and concentration of the
          toxicant applications.  Each application
          lasted for 5 hr, each increment was 1 hr.
          Interval 2=0-4= 171 - 216 hr.  Data
          were not recorded during Interval 1 which
          was the 24 hr prior to hour 0 (see text
          for explanation) ..... 	   27

   15     Average ventilatory rate for a bluegill
          during Intervall 1 and the first three
          days of Interval 2.  The fish was placed
          in the sensing chamber during the first
          hour	   27
                               : x

-------
                            TABLES

Number                                                      Page

   1      Concentrations of heavy metals used in
          ventilatory studies 	   24

   2      Example of partitioning of ventilatory
          rate data for analysis	   31

   3      Results of combined toxicity tests  	   33

   4      Results of chi squared tests for goodness
          of fit to the normal distribution	38

   5      Cochran's C Statistic 	   39

   6      Results of one-way ANOVA  	   40

   7      Results of Duncan's Multiple Range test at
          the 95 percent level of significance	41
                          APPENDIX A

  A-l     Results of chemical analyses given as the
          average of the five analyses for each fish
          bioassay chamber	 .   51

  A-2     Toxicant concentrations and survival during
          toxicity tests  	   54
                               XI

-------
                         ACKNOWLEDGMENTS

     This study would not have been possible without the
cooperation and assistance of the following people:  J. H. Alls,
D. D. Donald, B. F. Higginbotham, D. Hooley, J. D. Landers, Jr.,
G. L. Nunn, W. J. Showalter, and V. A. Wagner.
                              xii

-------
                             SECTION  1

                           INTRODUCTION


      The  increasing volume and  complexity  of  industrial,  agricul-
 tural ,  and  other  wastewater effluents in the  modern world make
 imperative  the  development of rapid  methods for  early  detection of
 developing-toxic  situations.  Use  of a  living organism as a    	-:
 generalized monitor capable of  responding  to  the entire  toxicant
 load  of its environment  has been suggested by numerous authors
 (see  Cairns and van der  Schalie, in  press, for a review).  A
 primary requirement for  such a  living monitor would be a detectable
.response  to sublethal  conditions as  a warning of the developing   •.
 lethal  situation.   Many  physiological and  behavioral parameters
 might be  used as  indicators of  stress situations.  One that shows
 considerable promise and has received attention  by many  workers
 is  the  rate at  which a fish passes water across  its gills during
 "breathing" activity (ventilation).

      The  ventilatory rates of fishes have  been used as an indica-
 tor of  stress for many years, and  numerous methods have  been
 employed  for observing the ventilatory  process.   The oldest and
 simplest  method is visual counting of the  opercular movements, a
 method  that has proven useful for  short-term  studies for many
 years (Belding, 1929;  Meuwis and Heuts, 1957;  Chizar et  al.,
 1972; Gee et al.,  1977;  Sorensen and Burkett,  1978).

      Many workers have measured ventilatory rates with surgically
 attached  or implanted  sensors of various types (Saunders, 1961;
 Pasztor and Kleerekoper,  1962;  Schaumburg  et  al., 1967;  Cairns
 et  al., 1970; Davis, 1972; Sparks  et al.,  1972;  Young,  1971;
 Hughs,  1975;  Lunn et al.,  1975; Sellers et al.,  1975;  Barnett
 and Toews,  1978).   These attachments may contribute to the stress
 of  the  fish,  and  the necessity  of  making these attachments,
 complicated or  not, to each individual  specimen  precludes the use
 of  such methods for gathering large, long-term data sets from
 large numbers of  individuals and species.  Such  data,sets are an
 absolute  necessity if  any validity is to be attached to  pre-
 dictions  made with such  information.

      The  most recently developed method for determining  the
 ventilatory rates of fishes is  also  the one that lends itself
 best  to the accumulation of a large  data base from numerous
 individuals and species.   This  method detects, by submerged,
 noncontact  electrodes  in aquaria,  the weak electrical  signals

-------
produced by the movement of the muscles in the branchial region
as a fish ventilates the gills during "breathing" activity
(Figures 1 and 2).

     Possibly the earliest workers to use such noncontact sensing
were Kleerekoper and Sibakin (1956) who investigated the weak
electrical signals produced by the lamprey; however, their subject
was physically restrained in a plastic tube.  Camougis (I960)
used noncontact electrodes to sense the bioelectrical potentials
of fish and other aquatic animals that were unrestrained in an
aquarium, and thus presumably, were less stressed than when
restrained.  Probably the earliest use of this type of apparatus
to describe the response of freshwater fish to environmental
variables was that of Roberts (1964) who studied the effects of
temperature and photoperiod on Lepomis gibbosus.  This type of
apparatus has been used more recently to investigate ventilatory
response to various environmental contaminants (Spoor et al.,
1971; Drummond et al., 1973; Drummond et al., 1974; Thomas and
Rice, 1975; Drummond and Carlson, 1977; Carlson and Drummond,
1978; Maki, 1979; Thomas and Rice, 1979).  While some of these
investigators reported the effects on ventilatory rates, the
majority emphasized the cough or "gill purge" frequencies with
little emphasis on actual ventilatory rates; or'their changes.
Most noncontact studies., .mentioned thus far'have ..relied -on the
analysis of short intermittent segments of strip chart recordings
which leave a large amount of the ventilatory behavior undescribed.

     Recent technological advances in.the field of electronics
have made relatively low-cost, dependable minicpmputers, micro-
processors, or other electronic recording devices economically
feasible for laboratory use.  These devices are easily adaptable
for the continuous accumulation of fish ventilatory data and make
possible the establishment of the*large'data bases necessary for
statistical predictions and valid long-term environmental moni-
toring.  Results from such systems have, been reported by Morgan
and Kuhn (1974), Morgan (1977) y-Westlake and van der Schalie
C1977), Thompson et al. (1978), Morgan et al. (1979), van der
Schalie et al. (1979), and Cairns and Thompson (in press).

     Most toxicity data are based on conventional tests in which
constant toxicant concentrations are maintained and in which the
time to death of the test organism^is; the measured parameter.
However, fluctuating conditions are the rule in the "real world"
and constancy the exception.  In order to develop a living,
continually sensing, water quality monitor, controlled "real
world" conditions must be simulated in the laboratory for deter-
mining predictable stress responses.  The non-stressed behavior
of the test organism must first be described for comparison.

     The present study was instigated to investigate and describe
the ventilatory response of the bluegill (Lepomis macrochirus)
to simulated "real world" conditions  (that  is, to fluctuating,

-------
Figure 1.  A bluegill in a ventilatory sensing chamber.  Water
flows into the receiving chamber (A), through the sensing chamber
(B), and out the drain (C).  The electrodes are the S-shaped
wires at either end of the sensing chamber.  The signal is carried
to an amplifier (AM).
           5
                            10
                                SECONDS
 Figure  2.   An example  of  the  amplified  ventilatory  signal  of  a
 bluegill ventilating at approximately 20  times  per  minute.

-------
complex, sublethal toxicant concentrations) and to accumulate a
large enough data base to allow valid statistical characteriza-
tion and analyses.  The..toxicants used..are all common components
of industrial effluents.   Standard" 96-h'r-median lethal concentra-
tions under conditions in this laboratory were established for
all the toxicants used in order to relate to conventional toxicity
data and determine the concentrations to be used in the ventila-
tory studies.

-------
                            SECTION 2

                           CONCLUSIONS

1.   Ninety-six hour median lethal concentrations for divalent
     zinc,  copper,  and nickel,  and for hexavalent chromium were
     determined to be 3.2, 1.0, 21.2, and 132.9 mg/1, respectively.

2.   At the concentrations tested, the toxicities of mixtures of ........
     zinc and copper and of zinc, copper, nickel, and chromium
   .  were determined to be additive.

3.   The ventilatory response of the bluegill (Lepomis macrochirus)
     was shown to be measurably affected by sublethal concentra-..
     tions of zinc, copper, and the complex mixtures indicated
     above.  This effect can be used as an indication of stress
     in an automated biological monitor.

4.   Ventilatory data meet the requirements for parametric
     statistical analyses when comparisons are between fish and,
     thus,  when sample size is large enough, can be analyzed by
   .  conventional methods.

5.   The bluegill displays two sensitive ventilatory responses to
     sublethal amounts of heavy metals: the ventilatory rate
     increases and the amplitude of the ventilatory signal
     decreases.  Both of these changes are statistically detectable
     and are reversible on removal of the toxicant within the
     limits of this study.

6.   Under the conditions of these tests, the addition of the
     heavy metal compounds could account for only a small portion
     of the reduction in signal amplitude indicating that the
     observed reduction was primarily due to a real change in the
     ventilatory behavior of the fish.

7.   No statistical differences could be demonstrated in the
     average ventilatory response with respect to the pattern in
     which the toxicant was applied.  However, graphic results
     indicated that the fish tend to respond to each successive
     toxicant application in a similar manner, even when only 5
     hours recovery time was allowed.
                                5

-------
                            SECTION 3

                      MATERIALS AND METHODS

TEST SPECIMENS

     The test organism was the bluegill, Lepomis macrpchirus,.
All specimens used during this study were obtained from Osage
Catfisheries, Inc. of Osage Beach, MO and were maintained in
450-& continuous flow aquaria.  The light regimen was continuous
light, and the temperature was maintained at a constant 22.5+l°C.
These conditions were maintained for both acute toxicity.and
ventilatory studies.                    .             .    .

Acclimation and Health

     The fish were acclimated to these holding conditions 'for at
least 2 weeks before being used for tbxi city..,test's and for 4 .to
6 weeks, and often as long as six months or more, before being
used for ventilatory studies.                          ,

     All fish were free of any visible disease-(mortality
-------
Proportional Diluter

     A solenoid valve controlled proportional diluter system
similar to that described by Hendricks et al. (1977) and Peltier
(1978) was used for the toxicity tests.  In this system, six
fish holding chambers of 6.4 i capacity received 233 ml of solu-
tion every 3 min providing a flow rate of 4.66 I per hour
(Figure 3).  This rate was equivalent to 17.5 chamber volumes
during each 24-hr period.

     The head boxes, proportioning boxes, mixing chambers, and
fish holding chambers were constructed of 0.64 cm plexiglas with
methylene chloride fused joints.  Delivery tubes and standpipes
were of teflon, and fittings were of glass.  The toxicant
reservoir was polyethylene (Figure 3).  The diluter was scrubbed...
out and then flushed with a solution of ethylenediamine tetraacetic
acid (EDTA) between toxicity tests.

     The toxicant solutions were delivered to the mixing head box
by a variable speed peristaltic pump.  The concentration of
toxicant in the fish chambers could be changed between tests by
one or a combination of the following four methods: the concen-
tration of the toxicant in the toxicant reservoir could be
changed, the flow rates of the diluent water or of the toxicant
solution could be changed, or the percentage of toxicant could be
adjusted.  The latter could be accomplished by changing the
height of the adjustable standpipes in each compartment of the
proportioning boxes (Figure 3).

Protocol for Toxicity Tests

     Ten fish were placed in each of the fish chambers at the
start of each toxicity test.  The number of individuals surviving
in each chamber was recorded after 1, 2, 4, 8, 12, and 24 hr and
every 24 hr thereafter.  All dead fish were removed at each
observation.  Death was defined as the loss of equilibrium.  Fish
that could not respond to prodding with a blunt rod by righting
themselves were considered dead.

Probit Analysis

     Probit analysis as described by Finney (1971) was performed
to determine the 96-hr median lethal concentration (LC50) values
and their 95 percent confidence limits.  A computer program
available on SAS79 was used to carry out this analysis (SAS
Institute, 1979).

     Two toxicity tests were run with each toxicant solution.
Toxicant concentrations were adjusted slightly when necessary
for the second test to allow for better resolution at concentra-
tions near the LC50 since probit analyses are more meaningful

-------
             OECHLORINATED
             TAP WATER 22*C
              HEAD BOXES
              'SUSS
                                                        TOXICANT
                                                        OEUVERY
                                                         PUMP
              MIXING CHAMBERS
                FISH
              CHAMBERS
                     i     22    3m    4m    sn     i
                   nOKOUENT WTOMCANT BKTOXCANT 3TBn»X»NT5BCTOqC4NT OOXTOOCM4T
                                                     TOXICANT RESEHVOR
Figure,3.   The solenoid controlled,  continuous flow,  proportional
diluter.   The time delay relay simultaneously opens  the valves  on
the head boxes and closes those on  the proportioning boxes allowing
the proportioning boxes to fill with diluent and  toxicant solutions.
When the level of .the  liquid i-n the diluent proportioning box
reaches the'"tbp in compartment 5, a float. yaJ.ve is tripped and
the solenoid valves  are returned to their normal  state, delivering ..
toxicant and diluent first to the mixing chambers and ultimately
to the fish chambers.   This also resets the time  delay relay,
and the cycle is repeated.

-------
with mortalities other than 100 percent or 0 percent.  The results
of successive tests with the same toxicants were in fairly close
agreement, and the results of tests using the same toxicant were
pooled to increase sample size.  The probit analysis was run on
the pooled data.

Joint Toxicity

     Since organisms may be more sensitive to some toxicants than
to others, a method of equating the toxicity of the various com-
ponents of a mixture was needed.  Sprague (1973) recommended that
the concentration of a toxicant in a mixture be reported as a
proportion of the LC50 as determined singly.  In this method, the
sum..of -the ratios for all components of a toxicant mixture is
determined.  If this total is 1.0, the toxicity of the mixture is '
considered to be additive.  A total of less than 1.0 is considered
to be synergistic (greater than additive), and a total greater
than 1.0 is considered to represent a mixture with antagonistic
(less than additive) toxicity.  Marking (1977) used this concept
as a basis for the development of a linear additive index in which
values of zero indicate additive toxicity, positive values indicate
synergistic toxicity, and negative values indicate antagonistic
mixtures.  Marking's additive index (MAI) has the advantage of
providing a method for establishing critical limits for values
that are near zero.

     MAI can be calculated using the equation:



                        \ Ami/Ai  - S                             .


where A  ./A. = toxic unit, i = the total number of toxic components
in the mixture, A . = LC50 for a toxicant in the mixture, A. = LC50
for that toxicant as determined individually, and S = the sum of
the biological, effects or the sum of the toxic units.  Then MAI =
(1/S) - 1.0 if S £ 1.0, and MAI = 1.0-SifS>^1.0.

     The significance of the deviation from zero can be determined
by substituting values of the 95 percent confidence limits for
the LC50 values of both A. and A  . into the formula for MAI to
establish a range for MAI.  The values that yield the greatest
deviation from MAI are used to establish the range.  If the range
includes zero, the deviation is not considered to be'significant.

WATER CHARACTERISTICS

     Dechlorinated tap' water from a single source was used as
diluent water for all toxicity tests and ventilatory studies and
to supply the holding tanks.

-------
Temperature

     Water.was maintained .at.a.constant, temperature with a Therma-
treck 34-E temperature controller.  A malfunction of the tempera-
ture controller during the first toxicity test using copper
resulted in a temperature of 19 C +_ 2.5 C.  In all other toxicity
tests and all ventilatory studies, the temperature was maintained
at 22.5°C +_ 1°C.  A Tempscribe remote bulb temperature recorder
was used to monitor the temperature of the diluent water con-
tinuously.  Additionally, daily temperature measurements were
taken from each fish chamber of the proportional diluter when
toxicity tests were being run.

Water. Chemistry

     Dissolved oxygen, conductivity, pH, alkalinity, and total
hardness were measured in each of tjie six fish chambers at the
beginning of every toxicant test and once -during each subsequent
24-hr period for the duration of a test... Consequently, five
measurements were taken from each chamber during ;the tests.  Since
the diluent water for both toxicity testing and ventilatory
studies 'was from the. same source, chemical data.::were not measured
for the ventilatory chambers;  The chemical values of the diluent
water ic the ventilatory ^studies. Is assumed•.•• to be the same as
those measured from chamber 1 (control) of the toxicity tests.

     Dissolved oxygen was measured with a YSI Model 54 A oxygen
meter.  Conductivity in micromhos/em was. measured with .a Bepkman
Model SM-210 conductivity meter modified to read in the 0 - 500
micromho range.  When concentrations of toxicant .were high enough
to cause values greater than 500, the samples were diluted with
distilled water (conductivity = 0 micromho/cm).  A Corning Model
7 pH meter was used to determine the pH.  Alkalinity and total
hardness were determined tltrimetrically according to methods
described by American Publi.e..J3ea.lth Ass.pciatlon (APHA) et al.
(1976).  Total-'hardness could hot be detefml'he'd in thi;sv.manner
when nickel was present and was determined instead by measuring
the calcium.and magnesium/concentrations"by atomic absorption
spectrophotometry and calculating total hardness by;the methods
given by APHA et al.- (1976).  -  . ' "...                "'  r      .

   ' •  The dissolved oxygen concentration was.maintained kt'Tevels
above 6.6 mg/& in all fish chambers throughout the toxicity '     -.
testing.  Dissolved oxygen during ventilatory studies was measured
only at the end of a test and averaged  7.7 mg/£.

     Conductivity  of the diluent water ranged from 100 to 145
micromhos/cm, pH from 6.7 to 7.6, alkalinity from 22 to 42 mg/Z,
and total hardness from 16 to 82 mg/Jl.  High concentrations of
toxicants, especially of hexavalent chromium caused the con-
ductivity to go as high as 3352 micromhos/cm, the pH to reach 8.3,
and the alkalinity to reach 564 mg/fc.  The average values for

                               10

-------
each fish chamber during each toxicity test are given in Appendix
A (Table A-l).

     A Wallace and Tiernan Amperometric Titrator was used to
periodically monitor the total chlorine content of the diluent
water.  The total chlorine level remained between 0.01 and 0.05
mg/1 throughout the course of these studies.

Toxicant Measurement

     Toxicant concentrations, as mg/Jl of the metal ion, were
determined with a Perkin-Elmer model 460 atomic absorption
spectrophotometer.  Analyses were done using an air-acetylene
flame except when both nickel and chromium were present.  In ........
that case a nitrous oxide-acetylene flame was used for the
determination of the concentration of the hexavalent chromium     :
since nickel blocks the detection of chromium.  Water samples
for determination of the concentration of the toxicant were
acidified with concentrated nitric acid, sealed with parafilm, ..:,.-•
and held until the end of a test before being analyzed.

     During toxicity testing, five water samples were taken as
indicated for water chemistry.  During ventilatory studies, water
samples were collected below the "drain without disturbing the
fish.  These samples were taken when appropriate depending on the
toxicant application pattern.  This was arranged so at least four
samples were taken at each toxicant level, and at least one sample
before and one after the toxicant was applied.  While each sample
was analyzed separately the average concentration for each fish
chamber or ventilatory sensing chamber was used for final analyses.

Toxicant Solutions

     The toxicant sources were reagent grade zinc sulfate (ZnSCK-
7H20), cupric chloride (CuCl2-2H20), nickelous chloride (NiCl2-
6H20), and potassium chromate (K2Cr04).  To provide divalent zinc,
copper, nickel, and hexavalent chromium these compounds were
dissolved in distilled water.  When chromium was one of the metals
in a toxicity test, an airstone was placed in the mixing head box
of the proportional diluter to facilitate mixing.  The chromium
solution was so concentrated that a density gradient between the
diluent water and the chromium solution formed and resulted in a
very concentrated layer of toxicant solution in the bottom of the
mixing box.  However, no precipitate was observed in'the mixing
head box.

     When determining the toxicity of complex solutions, it is
desirable to control the ratios of the toxic strength of the
various toxicants so that one component is not the major contri-
butor of toxicity.  Tsai and McKee (1980) defined the toxicity
ratio in the following manner:
                               11

-------
     Toxicity ratio = *mI/\ :  &m2/A2 "" Ami/A2               (2)
where the definitions of the terms are the same as for equation (1)

     Due to interactions between the dissociation products of the
various chemical compounds and between these products and the
diluent water which resulted in various precipitates, separate
solutions of each compounds for use in joint toxicity tests were
prepared.  Each was then pumped separately into the toxicant head
box where mixing occurred (Figure 3).  Because of these complica-
tions, the desired toxicity ratio of 1:1 was achieved only for
MixA (Zn++, Cu++).  The toxicity ratio ultimately achieved for
MixB (Zn++, Cu++, Ni++, Cr+6) was approximately 1:1.8:1.7:1.6.
However, since the primary purpose of these toxicity tests was
to indicate concentrations in the sublethal range for use in
ventilatory studies, these results were accepted and no further
toxicity tests were carried out.

VENTILATORY STUDIES                  .

     The electrical action potential of muscle movement in the
branchial region of a fish during ventilatory or: "breathing11
activity produces an electronic signal. -'A system for recording
this signal continuously and-for applying toxicants to the fish
in varying, pre-selected patterns was constructed.  •         ., ;

Sensing Chambers                 •'•"'.

     A fish ventilatory sensing chamber with a total of 16  :
individual 700 ml capacity monitoring compartments was constructed
of 6.35 mm transparent plexiglas.  The ends and dividing baffles
were of 3.18 mm transparent plexiglas, and the sides of each
individual compartment were of 3.18 mm opaque white plexiglas
(Figures 4 and 5).  All joints .w.ere fused.wi.th methylene chloride.

Electrodes                            :..    	''  .   •-,..,.

     The submerged electrodes were of ...type 304 stainless steel
wire and were attached with silicone aquarium cement to t.he.-baffles
at either end of the monitoring compartment.  The electrodes were
on the side .of the dividing baffles opposite the fish:to'prevent
contact between fish and electrode; 'Each eleetrode,,was 'formed   :.V
in such a way that the wire never passed through the flow of
water from any of the perforations in the dividing baffles
(Figure 4).  The electrode wire was insulated from just below the
water line, where the insulation was sealed to the wire with
silicone cement to prevent water intrusion, to the barrier strip
on the outside of the tank.  The only exposed wire was below the
water surface and at the barrier strip connections.

Isolation Chamber

     Since fish respond to visual stimuli and to vibrations in

                                12

-------
                            FISH HOLDING

                             ^ELECTRODES
                       DRAIN
  TO AMPLIFIER
Figure. 4..  An individual  ventilatory sensing chamber.  The    	
dimensions of the fish holding chamber are:   length, 12 cm; width',
4 cm; height, 10 cm;  and  water depth,  6 cm.
Figure 5.  Cutaway  view of the plywood isolation chamber to show
the composite  ventilatory sensing chambers.
                                13

-------
a manner similar to their response to chemical stimuli, the
entire fish holding chamber was enclosed in a plywood isolation
chamber.  The supports of this enclosure were anchored in sand
filled containers to protect against vibration (Figures 5 and 6).

Water Delivery

     Diluent water was delivered by gravity ,flow to three
plexiglas mixing chambers where the diluent could be mixed with
various toxicant solutions as desired (Figure 6).  The water then
was passed into delivery chambers which were also of plexiglas
and equipped with 1 mm diameter standpipes. A constant head
pressure was maintained over these standpipes'by delivery in
excess with the excess drained through an adjustable standpipe.
This allowed delivery of water or water and toxicant at a constant
rate and concentration to each sensing chamber (Figure 6).

Toxicant Delivery

     An Intergral Masterflex, variable speed, peristaltic pump
drive with remote control capability-.was used to introduce the
toxicant into the mixing chambers (Figure 6).- -As many as nine
pump heads could be used simultaneously for delivery of different
toxicant solutions tq.: each mixing chamber or for producing
various combinations in the mixing chambers.  The toxicant solu-
tions were pumped through 1166 mm inside -diameter silicone tubing.
One of the three mixing chambers was used-'for control and," thus,
never received toxicant solution.  Duplicate toxicant reservoirs,
pump heads, mixing chamberst and water delivery chambers (Figure
6) allowed for replicates of the treatments during each ventila-
tory test.  Two different treatments during a single test could
have been carried out using this apparatus.  Between ventilatory
runs, the system was rinsed with a solution of EDTA.  The mini-
computer used a digital to analog converter to deliver a pre-
selected voltage to the pump drive.  A computer-pump interface
was used to convert the output voltage from, the computer to the
optimum range for use by the pump.  The.computer output voltage
could be changed by the computer at pre-selected 1-hr -intervals
to simulate fluctuating "spills" or releases.  The concentration  '
of toxicant in the individual sensing-compartments reached
maximum concentration in 15 - 30 minutes and dropped -to backgroud
level, at approximately the same" rate,'depending'on"the .flow, rate
through the compartment (Figures 7 and 8>. ,.,.,.,, :

Amplifiers

     An amplifier designed for use with fish ventilatory signals
has been described in detail elsewhere (Gruber et al., 1977).
Sixteen of these high-gain noise immune amplifiers, one for each
of the sensing compartments, were constructed for this project.
The total gain of this amplifier is variable from 29,200 to
321,000 times.  Due to individual variability among test subjects,

                               14

-------

Figure 6.  The water delivery and data accumulation system.
                               15

-------
                       3.0
                     ~2.5'
                     o>
                     J
                     Z
                     22.0
                     5
                     a:
                       1.3
                     8
                       1.0
                      0.0
                                          137'ml/miii
                                          108 trt/min.
                         0      IS     30     45    . ..6O
                        MINUTES AFTER TOXICANT FLOW.STARTS
Figure 7.   The rate of  increa.se of the concentration in  the
ventilatory sensing chambers. •'  ••'
                       3.On
                     z
                     S2.OH
                     <
                       i.s-
                     o
                     o
                     -f 1.0-
                     2o.5-
                       0.0
                                         •;. .I37ml/min.
                                           O8ml/mia :
                         0      15     30     45      60
                         MINUTES AFTER TOXICANT FLOW STOPS
 Figure  8.  The rate  of decrease of the  concentration  in the
 ventilatory  sensing  chambers.
                                    16

-------
each amplifier was adjusted during the acclimation period to pro-
duce a ventilatory signal in the 4 to 8 volt range; consequently,
comparisons of absolute signal amplitude between fish are rela-
tive.

     Design modifications that increase the maximum range of this
unit for use in marine situations have been described recently
by Cairns et al. (1980).

Data Accumulation

     Each of the 16 ventilatory monitoring compartments was
connected to an amplifier.  These in turn were interfaced to a
PDP-8E minicomputer through a 16-channel multiplexed analog to ...,
digital converter (Figure 6).

     Ventilatory counts and an average scaling factor for each  •
fish were accumulated for 15-min periods.  The scaling factor is
a value from which an approximation of the average signal ampli-
tude can be recovered.  - At the end of each of the 15-min periods,
the accumulated data were printed on a Decwriter II computer
terminal.  At 30-min intervals, the two preceeding 15-min data
sets were stored on magnetic tape cassettes by a Decassette TU60
tape drive.

     At the end of each ventilatory run, the minicomputer was
connected by telephone line to an IBM/370-1580, and data stored
on the cassettes were transcribed onto a Conversational Monitoring
System (CMS; an interactive programming system for the IBM
System/370). Transcription took approximately 1 hr.  The mini-
computer could then be reset and another ventilatory study started
as desired.

Computer Program

     The ventilatory signal is often bimodal (Thompson, et al.,
1978; Gruber et al., 1979).  This bimodality is an artifact of
the amplifier design and is more pronounced at slower rates
(Thompson et al., 1978).  The signal has been visually correlated
with the ventilatory activity of L. macrochirus, and each waveform,
bimodal or not, represents one complete ventilatory cycle (Gruber
et al., 1979).

     The computer program developed for this project'is capable
of rejecting this artifact when it is present and of accumulating
the total ventilatory counts as well as an approximation of the
average amplitude of the signal.  This program also uses a soft-
ware controlled variable threshold to reject low amplitude noise.
Signals with less than 1 volt peak-to-peak amplitude were con-
sidered to be noise.
                               17

-------
     This program used the first derivative of the ventilatory
signal for determining the number of cycles or waveforms in a
reporting period.  In each cycle there:.was at least one region
in which a relatively large change in the signal voltage resulted
in a large peak in the derivative signal at the point of maximum
change.  A software Schmidt trigger was used to reduce the effect
of noise on the signal.  This was done by establishing symmetrical
positive and negative thresholds for the derivative signal.  When
the derivative signal exceeded the negative threshold, the count
logic was cleared.  When a following positive signal exceeded the
positive threshold, the logic was clocked and a count registered.
Consecutive pulses of the same polarity did not produce counts
unless there was a pulse of opposite polarity occurring between
them.  Examples of the computer produced derivative signal are
shown in Figures 9 and 10.  The sharp negative pulse indicates
that a count has been registered and further counts will be
inhibited until the logic is reset by a negative pulse.

     The similarity in shape of the electronic artifact to some
types of gill purge signals (van der Schalie.,.,:1980) precludes the
use of this program for detecting that response.  Other types of
gill purges appear a,s: .sigrialis,'^pf.. .higher amplitude than the
associated ventilatory signals.  These, as well as any body move-
ments that produce signals^ meeting the above criteria, would be
included as ventilatory. signals-.  .

     The amplitude of the input signal was variable over time and
between individual fish which limited the usefulness of the fixed
thresholds.  To alleviate this problem, an automatic gain control
was used that kept track of the amplitude of the incoming signal
for 2 sec and adjusted the scaling factor for each succeeding
signal either up or down to result in a predetermined peak signal
level.  A maximum scaling factor'prevented signals with amplitudes
below a specified level (1 volt peak-to-peak) from being amplified
beyond the threshold and,.thus,..producing false counts.  This was,
in effect, a variable threshold which allowed a more accurate
interpretation, of the ventilatory events than was possible with
earlier biomonitoring programs that used'fixed thresholds with no
automatic gain control (Cairns et al. ,' 1977; van der Schalie,
1977).  With those programs, large reductions in amplitude would
have registered as a reduction.In rate.  With the present program,
real ventilatory--signaTs with amplitudes of less than 1 volt peak-
,to-peak would also be ignored, and a possible false redaction in  '
rate would ;be;indicated.

     The length of the time period during which the amplitude of
the signal was retained was variable and was determined by the
operator at the start of an experiment.  This time period should
be similar to the period of the wave form being observed.  If it
is too short, the smaller peaks may be amplified to the maximum
and result in false counts.  If too long, a real cycle of low
amplitude would be lost if a very large peak immediately preceeded

                               18

-------
                        -i—r
                                           —r-
                                           is
                              Seconds
Figure 9.  Strip chart recording of the ventilatory  signal  of  a
bluegill.  A.  The ventilatory signal as taken directly  from the
amplifier.  B.  The computer produced derivitive of 'the  signal
in A.  The very sharp negative peaks indicate that a count  has
been recorded.
                               19

-------
                                   10
                               Seconds
Figure 10.  Strip chart recording of the ventilatory signal of a
bluegill.  A.  The ventilatory signal as taken directly from
amplifier.  Note the increased bimodality with respect to Figure
9.  B.  The computer produced derivitive of the signal in A.
                               20

-------
it.  Two seconds worked well for bluegills and was the value used
for all of the ventilatory studies reported here.

     This program accumulates data with low error rates when in
use with input signals in the optimum range of 4 to 8 volts.
Normal cycles were missed less than 2 percent of the time, and
extra counts were added less than 1 percent of the time.  These
levels remained below 10 percent even when the signals were
irregular and of low amplitude.

Major Equipment Problems

DC Voltage Drift—

     The fish ventilatory signal is detected as a change from.,... .,.,„.,
zero voltage.  From time to time, the background voltage drifted.,;
from zero to a positive or negative DC voltage.  Detection of the
signal by the computer program is dependent on the signal breaking
both a positive and a negative threshold.  These thresholds are..
set at absolute values around zero, and, if the background voltage"
departs greatly from zero, the signal is lost.  When a sensing
chamber was empty or when a fish was between breaths, this
voltage should be zero in a properly adjusted system (Figure 2).
A 10,000 n loading resistor placed across the electrodes elimi-   '
nated most of the voltage drift.  However, as the equipment wore
out over time, the signal tended to drift from zero and data were
lost (Figure 11). The results from 10 fish during the first four
ventilatory studies were lost in this manner.

     This problem was caused by corrosion of the terminal barrier
strips which were the connectors between the electrodes and the
amplifiers.  These were inside the isolation chamber where the
atmosphere was very humid.  The problem was solved by replacing
all electrodes and barrier strips.  In practice, this is necessary
about every 18 months under conditions of continuous use.

Electronic Noise—

     While shielded cable was used for all wiring and proper
grounding techniques were used, a considerable problem with
electrical interference or "noise" still existed that could mask
the ventilatory signals.  The source of this noise appeared to
be ground currents flowing from the water supply to the drain.

     Water dripped into the sensing chambers from a height of 10
cm above the water surface provided a broken stream of water and
eliminated this noise (Figure 6).  The noise occurred only when
a continuous stream of water was present to carry ground currents.

Water Splash—

     The solution to the noise problem caused the lesser problem

                                21

-------
             24    48    72    96    120   144
                             TIME(HOURS)
168
192    216
 Figure  11.   The  average ventilatory rate-of a-'bluegill.  The
 arrows  indicate  points at which.the DC voltage drift caused a loss
: of-signal.  ' Replacing the' electrodes.' and -terminal barrier strips
 eliminated  this  problem.     .
                                22

-------
of splash from the dripping water sources.  Since controls and
treated fish occupied alternate compartments, sometimes toxicant
splashed into control compartments.  This was a minor problem,
and only one control fish was influenced by this; the data for
that fish was invalidated (#6, run 6).

Ventilatory Tests

     Twelve, ten-day ventilatory studies were carried out, three
with each of the four toxicants.  Each of the three toxicant
application patterns was included among these three tests, thus
all patterns were run with each toxicant (Table 1).

Toxicants-r

     The toxicants.used for the ventilatory studies were divalent;
zinc, divalent copper, MixA and MixB.  Neither divalent nickel
nor hexayalent chromium was tested separately.  All toxicants were
used at concentrations that were indicated as being sublethal
by the toxicity tests.  The amount of toxicant solution delivered
was approximately equivalent with respect to toxicity for each
test with a single toxicant;  however, this was not true between
toxicants, and, consequently, the most valid comparisons are
between patterns for each toxicant (Table 1).

Application Patterns—

     In the first application pattern, toxicant was delivered
to the sensing chambers at a constant level for 96 hr (Figure
12).  In the second pattern, toxicant was delivered in two
variable 20-hr applications separated by a 20-hr interval.  The
third pattern consisted of 8 variable 5-hr applications, each
separated by a 5-hr interval.  The 20-hour variable applications
were delivered in five 4-hr increments: the first at the same
concentrations as the constant 96-hr application, the second
at approximately twice.that concentration, and the third at three
times that concentration.  The concentrations then were reduced
in reverse order for the fourth and fifth increments, respec-
tively (Figure 13).  The 5-hr variable applications were incre-
mented in the same manner, but the duration of each increment
was only 1 hr (.Figure 14.).  A fourth pattern, zero, was designated
for statistical analyses.  Pattern zero was that of the control
fish which received no toxicant.

Test Protocol—

     Five of the 16 fish in each ventilatory test (3, 6, 9, 12,
15) were not exposed to toxicants and were used as controls.
The remaining 11 fish were exposed to various toxicants.

     Each ventilatory test was run for 102 days and consisted of
4 time intervals.  Ventilatory counts and average amplitudes were

                               23

-------
to
            TABLE  1.  CONCENTRATIONS OF HEAVY METALS USED IN VENTILATORY STUDIES

Toxicant
applications
One, 96 hr,
constant



Two, 20 hr,
variable



Toxicant
Zn++ Cu++ MixA
Run 1 Run 4 Run 7
Zn 2.5* Cu 0.5 Zn 0.6
Cu 0.2


Run 2 Run 5 Run 8
Zn 2.5, 5.7, 8.1 Cu 0.5, 1.0, 1.5 Zn 0.2, 0.4, 0.6



•
MixB
Run 10
Zn 0.3
Cu 0.1
Ni 2.5
Cr 25.5
Run 11
Zn 0.4, 0.
Cu 0.1, 0.
Ni 3.3, 5.
Cr 22.5,44.








8, 1.1
2, 0.2+
1, 8.0
9,58.7
    Eight,  5  hr,
     variable
      Run 3
Zn 2.9, 5.7,  7.7
    Run 6
Cu 0.5, 1.1, 1.3
      Run 9
Zn 0.6, 1.4, 1.8
CuO.2, 0.5, 0.6
   Run 12
Zn  0.4, 0.8, 1.1
Cu  0.1, 0.2, 6.2+
Ni  2.3, 5.3, 6.9
Cr 22.3,60.6,77.0
    *A11  concentrations are given as the average mg/1 of all fish treated during that run.

-------
                                                    FISH 7 TREATED

                                               26

                                                 FISH 14 TREATED
                                            FISH 9 CONTROL
                72   36   120   141  168  192  216

                   TIMEtHOURS)
Figure 12,  Toxicant  pattern 1.  Toxicant was delivered  to  fish
7 and fish 14 for  96  hr.   The shaded area on the fourth  vertical ,:
plane indicates  the shape,  duration, and concentration of the
application of the toxicant.  Fish 9 was a control and received.no
toxicant.  Interval 2 = 0 - 96 hr (X axis); Interval  3 = 96 - 192
hr; Interval 4 = 192  - 216 hr.  Data were not recorded during
Interval 1 which was  the 24 hr prior to hour 0 (see text for
explanation).
                                25

-------
                a jo
i6o;
120-
eo
        80

        40
                                   \ y  j •.•'••;' .  .  j
                                   Y7__  y^
                                    w« —. —^-^^—^ .x .
'VU^ju^/^^
                                                     S3

                                                      FISH 14 TPIEATED
    'as     ,.,..

     F\SH 13. TREATED
jus- '•
 OSH 9 CONTROL   .
         24   48   72   96   120   144  168  IB3  216
                    TIME (HOURS)
Figure  13.   Toxicant pattern 2.   Toxicant was delivered  to fish  ,
14 and  fish 13.   The shaded area on the.-fourth vertical  plane
indicates  the shape, duration,  and concentration of the  toxicant
applications.  Each application lasted for 20 hr; each increment
was 4 hr.   Interval 2 = 0- 96  hr. .(X axis); Interval -3 = 96 - 156-
hr; Interval 4 = 156-216 hr.   Data were not recorded during
Interval i.which was the 24 hr  prior to hour 0 (see text, for
explantion)
                                26

-------
                                                    FISH 16 TREATED
                                                FISH 8 TREATED
                                           216

                                            FISH 12 CONTROL
        24
            48
               72  36  120  144  168

                   TIME (HOURS)
                                       216
Figure 14.  Toxicant  pattern 3.  Toxicant was  delivered to fish
16 and fish 8.   The shaded area on the fourth  vertical plane
indicates the  shape,  duration, and concentration  of  the toxicant
applications.   Each application lasted for 5 hr,  each increment
was 1 hr.  Interval 2 = 0 - 96 hr (X axis); Interval 3 = 96 - 171
hr; Interval 4 = 171  - 216 hr.  Data were not  recorded during
Interval 1 which was  the 24 hr prior to hour 0 (see  text for
explanation).
                                27

-------
recorded only during the last 3 intervals, a total of 216 hr
(Figures 12, 13, and 14).

     Interval 1. acclimation—The fish were placed in the sensing
chambers and allowed to acclimate for 24 hr.  This 24-hr period
was usually sufficient time to allow stabilization of ventilatory
rates (Figure 15).  Amplifiers could be adjusted during this
interval to produce signals in the optimum range (4-8 volts
peak-to-peak) and, if necessary, the rate of water flow to the
chambers could be reduced to eliminate electronic noise.  No
adjustments were permitted during the remaining three intervals.

     Interval 2. pretreatment—The ventilatory rate and average
amplitude were accumulated for 15-min intervals for 96 hr.  No
toxicant was delivered during this interval which was to provide
control data for each individual fish.

     Interval 3.  treatment—Toxicant was delivered.  This
interval varied in duration depending on the pattern of applica-
tion and continued only until the toxicant delivery ceased for
the final time.  For pattern 1, this was 96 hr; for pattern 2,
60 hr; and for pattern 3, 75 hr.

     Interval 4. post-treatment—The recovery period was monitored
from the end of interval 3 to the end of the ventilatory study.
This was 24 hr for pattern 1; 60 hr for pattern 2; and 45 hr for
pattern 3.

     The duration of all intervals for pattern zero (controls)
was the same as for the treated fish in each ventilatory test,
respectively.

Data Treatment and Analysis

     Data from four consecutive, 15-min reporting periods were
summed.  The ventilatory counts were then divided by 60 to yield
the average counts per minute for each 1-hr period.  Since the
signal amplitude was reported as an average, the hourly sums
were divided by 4 to produce an average voltage for each 1-hr
period.  Nunn et al. (in press) have shown that such ventilatory
rate data are approximately normally distributed, and, thus,
standard parameteric analyses performed on them are statistically
valid.

     These averages then were used to calculate an interval
mean for each fish during each of intervals 2, 3, and 4.  Thus,
an average rate and amplitude were produced for each fish during
the pretreatment, treatment, and post-treatment intervals.  For
example, the internal average rate for fish 12 (Figure 14) during
the pretreatment interval was 31.0 cycles per minute; during the
treatment interval, the average was 31.2 cycles per minute; and


                               28

-------
                  n	1	1	1	1	1	1	1
                   12    24    36   48    60    72    84    96
                                TIME (HOURS)
Figure 15.  Average  ventilatory rate for a bluegill  during
Interval 1 and the first  three days of Interval 2.   The  fish was
placed in the sensing  chamber during the first hour.
                                29

-------
during the post-treatment interval 32.5 cycles per minute.  These
interval means were then used for statistical analyses.

     Since the amounts of toxicant delivered were equivalent for
each test with a single toxicant but were not equivalent between
toxicants, it was decided to analyze for differences between
patterns for each toxicant separately.  The interval means from
all ventilatory tests with a single toxicant were assembled into
data sets.  Each data set consisted of all the averages (interval
means) from one of the intervals (2, 3, or 4) for all three tests
with that toxicant.  For example, the first data set was assembled
from the pretreatment interval for all three runs that used Zn"1"1"
as the toxicant.  This was then partitioned into four cells
corresponding to the four patterns in which the toxicant was
applied (Table 2).  This was done for the ventilatory rates for
each time interval for each of the 4 toxicants resulting in 12
partitioned data sets.  In similar fashion, 12 data sets were
assembled for the average amplitude data.  Thus, each of the
various analyses were performed on a total of 24 data sets.  Even
though replicate treatment data were available, Treatments A and
B (Appendices B and C) were combined for analysis.

     A chi squared test for goodness of fit was performed to com-
pare the distribution of the interval means in each data set to
the normal distribution (Siegel, 1956).  Cochran's C statistic
was used to test for homogeneity of variances (Winer, 1971).
One-way analyses of variances for unequal sample sizes were then
used to test for differences due to the pattern in which toxicant
was applied,  A post-mortem Duncan's Multiple Range test was then
used to locate the major sources of difference (Winer, 1971).
Computer programs available commercially were used to carry out
the latter two analyses (SAS Institute, 1979).

     In addition to the application pattern and the toxicant, four
other independent variables that may have affected the ventila-
tory rate were measured for each individual fish.  The standard
length of each fish was recorded in millimeters.  The average
toxicity ratio was calculated by dividing the average concentra-
tion of toxicant by the LC50 of that toxicant and multiplying by
the total time that the toxicant was delivered.  When fluctuating
applications were used, this was done for each concentration level,
and the three values were summed.  For MixA and MixB, the values
for each individual component of the toxicant mixture were calcu-
lated and these values were summed.  To determine the total dose
delivered, the average toxicity ratio was multiplied by the aver-
age rate of flow through the chamber in liters per minute.  The
fourth variable was the maximum concentration reached (expressed
in terms of toxicity ratios) during the treatment interval.  For
pattern 1, this was the average concentration; for patterns 2 and
3, it was the average concentration reached at the peaks of the
variable applications.
                                30

-------
TABLE 2.  EXAMPLE OF PARTITIONING OF VENTILATORY RATE DATA FOR
        ,  ANALYSIS

        "Pattern 0Pattern 1Pattern 2Pattern 3
Toxicant  Interval  (control). (1 x 96 hr.). (2 x 20 hr)  (8 x 5 hr)
Zinc
run 1

Zinc
run 2

Zinc
run 3
4 means*   9 means
3 means
5 means
10 means
             10 means
Totals for         12 means    9 means     10 means     10 means
 analysis

 DC voltage drift resulted in loss of 3 fish from run 1; 3 from
 run 2; and 1 from run 3.


     A forward stepwise, multiple regression was then used to
indicate which, if any, of the independent variables, pattern,
standard length, average toxicity ratio, total dose, or maximum
concentration might have a significant effect on the ventilatory
rate during the treatment interval.  This analysis produces an
F ratio to indicate the significance of each variable.  An R-
Square value indicates the amount of the total variability
accounted for by the independent variables.

     Only the data for the treated individuals from the treatment
intervals were used for this analysis.  Separate analyses were
carried out for each of the four toxicants.
                               31

-------
                            SECTION 4

                     RESULTS AND DISCUSSION

TOXICITY TESTS

     The results of the 96-hr continuous flow toxicity tests are
presented in Table 3.

Zinc

     The acute toxicity of zinc to bluegills under conditions
similar to those in the present study (temperature 19 - 22.5°C in
soft water) has been reported by others.  All the studies used
either zinc sulfate or zinc chloride as the source of zinc.
Pickering and Henderson (1966) reported 96-hr LC50 values of 4.85
(C.L. 3.55 - 6.32), 5.37 (C.L. = 3.84 - 7.00), 5.46 (C.L. = 4.29 -
6.94), and 5.82 rng/A (C.L. = 4.61 - 7.48) in soft water at 25°C.
Cairns and Scheier (1957a) reported values between 2.86 and 3.78
tag/H in soft water at 18°C and 1.93 - 3.63 mg/X, at 30°C.  The work
of Cairns and Scheier (1957b) resulted in the highest reported
96-hr median lethal concentration (8.02 mg/2.) for these conditions.
The 96-hr LC50 of 3.21 mg/l (C.L. = 2.10 - 4.61) determined during
this study is comparable to the values reported in the literature.

Copper

     The acute toxicity of copper to bluegills under similar
conditions has also been reported using copper sulfate or cupric
chloride as a source of copper.  Pickering and Henderson (1966)
reported a 96-hr LC50 of 0.66 rag/I (C. L. = 0.52 - 0.85) at 25°C.
Trama (1954) reported a value of 0.74 mg/fc at 20°C; Patrick et
al. (1968) reported a value of 1.25 mg/Z at 18°C; and Benoit
(1975) reported a value of 1.1 mg/X, at 20°C.  The value of 1.0
mg/fc (C.L. = 0.9 - 1.2) determined here compares favorably with
those values.

Nickel                                       .

     Few reports of the acute toxicity of nickel to bluegills
under similar conditions are to be found in the literature.
Pickering and Henderson (1966) used nickelous chloride and reported
96-hr LCSOs of 5.18 mg/£ (C.L. = 2.02 - 6.88) and 5.36 mg/SL (C.L. =
4.4 - 6.74) at 25°C in water of 20 mg/X, total hardness.  In water
of 360 mg/X., the same authors reported an LC50 of 39.6 mg/X, (C.L. =
33.7 - 45.3) for nickel.
                               32

-------

Toxicant
Zn++
Cu++
Cr+6
K Zn++
a Cu++
Zn++
O3 ++
S Cr+6
TABLE 3. RESULTS
LC50
3.2
1.0
21.2
132.9
1.4
0.4
0.9
0.5
10.0
61.6
OF COMBINED
TOXICITY TESTS
95 percent C. L.
2.1
0.9
18.1
113.6
1.0
0.3
0.7
0.2
4.6
47.3
4.6
1.2
- 25.5
- 155.7
2.0
0.6
1.3
1.2
- 23.1
- 82.4

Total
fish
120
120
120
117*
119t
118* '

 Three and two fish, respectively, died of behavioral causes and
 were disregarded.
 Only 119 fish at start due to counting error.


     The LC50 of 21.2 mg/X, reported here falls between those  ....
two values.  Our tests were carried out in water with a total
hardness of 32 - 71 mg/i, values that are closer to the soft
water of Pickering and Henderson than to their hard water.
Pickering and Henderson used static tests and reported only the
calculated concentrations of Ni++ while the present tests were
continuous flow and the Ni++ concentrations were actually measured
five times during the course of the tests.  It is also of interest
that the measurements are for total nickel in acidified samples.

Chromium

     The acute toxicity of chromium (potassium chromate) has also
been reported in the literature.  Trama and Benoit (1960) reported
an LC50 of 170 mg/S, for hexavalent chromium in water with a total
hardness of 45 mg/fc.  The LC50 of 132.9 mg/5, (C.L. = 113.6 -
155.7) determined for hexavalent chromium during this study agrees
fairly well with those results.  Trama and Benoit used static
toxicity tests and apparently reported calculated concentrations
rather than measured concentrations for chromium.  Again, it is
well to remember that the concentrations given in the present
report are measures of the total chromium present in the acidified
sample.
                               33

-------
MixA

     The acute toxicity of zinc-copper., mixtures to bluegills has
apparently not been reported in the literature.

     An MAI of +0.218 was calculated with the data in Table 3.
The calculated index range (-0.644 to- +1.304) includes 0, and,
thus, we conclude that the toxicity of zinc-copper mixtures to
bluegills is additive under the test conditions.

MixB

     The acute toxicity of zinc-copper-nickel-chromium mixtures
to bluegills has not been reported in.the literature.

     MAI for this mixture is -0.738, which would seem to indicate
a tendency toward antagonistic toxicity.  However, the calculated
index range of -2.982 to +0.193 indicates that this is not a
significant departure from additive toxicity.  It should be kept
in mind that the actual toxicity ratio during these tests was
1:1.8:1.7:1.6, not the desired 1:1:1:1.  Before discounting the
possibility that these metals display reduced-:;toxicity in mixture,
additional toxicity tests should be carried out.

VENTILATORY STUDIES

     All living organisms display considerable natural variability
in morphological, physiological, and behavioral parameters.  Vari-1-
ability in the ventilatory behavior of fish was both expected and
found; this quality makes change and adaptability possible—and
makes analysis of biological data difficult.

     Strip chart figures illustrating variability in ventilatory
rates as well as in the shape and amplitude of the signal"are
included in Appendix D.  These figures, .especially those repre-
senting interval 2 (the pretreatment interval), show that the
analysis of short intermittent segments of such recordings make
characterizing ventilatory behavior difficult. .The figures in
Appendix D can be placed in perspective by comparison with
Figures B-5 and G.-5 which indicate the average rates and. amplitudes,
respectively, for. these fish«  The strip chart recordings from
all 4 fish were taken at the same points during run 5 (copper,
pattern 2).

Diurnal Variability

     It has been shown that natural diurnal light cycles have a
definite effect on the ventilatory behavior of fishes (Morgan et
al., 1979; van der Schalie, 1977).  By exposing the fish to
continuous 24-hr light, it was hoped that these effects would be
removed to simplify analysis of data.  The graphic data for the


                                34

-------
pretreatment intervals in Appendices B and C show no obviously
identifiable cycles.

Other Variability

     It can also be seen from Appendices B, C, and D that a great
deal of unexplained individual variability as well as variability
between individuals was apparent.   One example of variation that
can be explained can be noted in run 1 (Figs. B-l, C-l).  At
about hour 48, a malfunction of the temperature controller caused
the temperature to drop from 22°C to 16°C in less than 15 min and
remain there for approximately 6 hr at which point the tempera-
ture was returned to 22°C at the same rate.  A rather high peak
for 4 or 5 hr is apparent for many of the fish.  This corresponds
to the return to 22°C and is the result of both increased rates
and amplitudes.  The rather quick return to prestress levels and...
the maintenance of those levels among the controls indicate, that
there was no deleterious effect on the fish, at least for the .......
duration of the test.

General Trends

     A total of 55 control fish were monitored during this study.
The.average ventilatory rate for these fish during the pretreatment
interval was 31.7 counts per minute (range, 15.9 - 51.0).  During
the treatment interval, the average rate was 31.0 counts per
minute (15.9 - 45.8), and during the post-treatment interval, it
was 28.7 counts per minute (14.2 - 43.8).  The average rate for
the other 125 fish during the pretreatment period was 28.5 counts
per minute (.14.3 - 55.4).  It is possible that the small downward
trend in rate is related to physical condition over time since
the fish were not fed during the tests.  Van der Schalie (1977)
visually observed rates of 34.8 counts per minute (28.0 - 48.0)
during daytime observation.

     The average amplitude during pretreatment for the controls
was 5.5 volts (2.5 - 9.9); during the treatment interval, it was
6.0 volts (1.8 - 12.3); and during post-treatment, 5.4 volts
(1.8. - 14.2).  The average pretreatment value for the treated
fish was 5.8 volts (2.5 - 14.2).  No obvious trends are apparent.

Changes in Rate

     During the treatment interval, 26 control fish increased
their average rate with respect to the pretreatment interval
while 29 decreased.  The average change was a decrease of only
0.7 counts per minute (2.2 percent).  Average ventilatory rate
for 115 fish that received a toxicant treatment increased during
this interval while only 10 decreased.  The average change here
was an increase of 15.5 counts per minute (54.4 percent).
                               35

-------
     During the post-treatment interval, 39 of the control fish
decreased in average rate with respect to the treatment interval
while 16 showed an increase.  The average change was a decrease
of 2.3 counts per minute (7.4 percent). Of the 125 treated fish,
123 survived the treatment interval,  and 111 of these decreased
in average rate during the post-treatment interval while 12 showed
an increase.  The average change for these fish was a decrease of
13.1 counts per minute.  This was a 29.8 percent decrease with
respect to the treatment interval but 46.0 percent with respect
to the pretreatment interval.  This can be compared with the 54.4
percent increase during the treatment interval.

     The obvious conclusion is that sublethal, heavy metal stress
causes increased ventilation rates in bluegills while the removal
of that stress results in a return to rates near the pretreatment
levels.

Changes in Amplitude

    "During the treatment interval, 27 of the control fish
decreased in average amplitude with respect to the pretreatment
interval and 28 increased.  The average change was an increase
of 0.4 volts (7.3 percent).  Average amplitude decreased in 107
treated fish while 18 responded with an increase.  The average
change for these fish was a decrease of 2.1 volts (36.2 percent).

     During the post-treatment interval, 21 of the control fish
increased in average amplitude and 34.decreased.  The average
change was a decrease of 0.6 volts (10.0 percent) with respect
to the treatment interval.  Of the 123 surviving toxicant exposed
fish, 112 increased their average amplitude during this interval
while only 11 decreased.  The average change for these fish was
an increase of'2.4 volts.  This was a 64.9 percent increase with
respect to the treatment interval and a 41.4 percent increase with
respect to the pretreatment interval.  This value can be com-
pared with the 36.2 percent decrease during the treatment interval.

     It is apparent from this that the application of these
toxicants caused a marked decrease in the average amplitude of
the ventilatory signals.  It is further apparent that recovery
to near pretreatment levels is indicated with the cessation of
toxicant application.

Toxicant Effects on the System

     It is known that large increases in conductivity or, corre-
spondingly, reductions in resistance can greatly reduce the
amplitude of the ventilatory signal.  Cairns et al. (1980) reported
that 29  /oo sea water reduced the signal amplitude so much that
a considerable increase in amplification was necessary in order
to adapt this system to marine conditions.  This raises a question
with regard to the source of the reduction in signal amplitude

                               36

-------
observed when toxicant was applied.  Since any addition of dis-
solved salts can reduce the electrical resistance of the water
between the electrodes (see conductivity values, Table A-l), it
is possible that the observed change in amplitude is due to the
effect of the toxicant on the electrical resistance of the system
and not to an effect on the fish.

     It is possible to measure the total resistance of a ventila-
tory sensing chamber by measuring the resistance when the loading
resistor is removed and a volt-ohm meter is attached in its
place.  The total resistance for a normally operating ventilatory
sensing chamber is then the inverse of the inverse of the sum of
the measured resistance and the 10,0000 loading resistor.  The
total resistance was measured for a chamber filled with diluent
water, various concentrations of the individual toxicants up to
120 mg/Jl (dissolved in diluent), and diluent water with a fish
in place.  The fish itself had no measurable effect on the total
resistance, but the dissolved salt content did.  The values ranged
from a high of 8389fi with diluent or diluent and fish to a low
of 714312 with a solution of 120 mg/fc Cu++.

     It was then possible to simulate the change in total system
resistance that would occur with the addition of various concen-
trations of toxicant by changing the values of the loading resistor
while only diluent was passed through the chamber.  Preliminary
data from an acclimated fish indicated that in order to reduce
the signal amplitude by 24.8 percent it is necessary to reduce
the total system resistance to approximately 6000J2.  The amplitude
reduction indicated for the measured solutions (100 - 200 mg/fc)
was 12-13 percent.  The maximum concentration delivered during
this study, test 12, would cause an amplitude reduction of less
than 10 percent.  Thus, the average amplitude reduction of 36.2
percent would be due primarily to the effects of the toxicant on
the fish, not on the electronic system.  It is also significant
to note that only in tests 11 and 12 did the toxicant concentration
reach levels that would depress the amplitude by more than 2
percent.

Distribution of Interval Means

     The distribution of each of the 12 sets of interval means
for the ventilatory rates was compared with the normal distribu-
tion using a chi squared goodness of fit test.  The results of
these analyses are presented in Table 4.  The null hypothesis
that no difference exists could be accepted at a traditional
alpha level of 0.05 for nine of these distributions.  Two of the
remaining three had probabilities for acceptance of null greater
than 0.01 while only one had a probability for acceptance of less
than 0.01.

     The distribution of each of the 12 sets of interval means
for average amplitude was also compared with the normal distribu-


                               37

-------
  TABLE 4.   RESULTS OF CHI SQUARED TESTS FOR GOODNESS OF FIT TO
            THE NORMAL DISTRIBUTION ...

Interval
Toxicant
Zinc
Zinc
Copper
Copper
MixA
MixA
MixB
MixB
Data set
Rate
Amplitude
Rate
Amplitude
Rate
Amplitude
Rate
Amplitude
Pretreatment. .
0.80 -
0.20 -
0.10 -
0.05 -
0.01 -
0.05 -
0.10 -
0.50 .-.
0.90*
0.30
0.30
0.10
0.02
0.10
0.20
0.70
. Treatment. .
.0.20 -
0.20 -
0.05 -
0.02 -
0.001-
0.001>
0.70 -
0.02 -
0.30
0.30
0.10
0.05
0.01

0.80
0.05
Post.-.treatment
0.70
0.30
0.20
0.70
0.10
0.02
0.01
0.30
-0.80
-0.50
-0.30
-0.80
-0.20
-0.05
-0.02
-0.50

 Probability for obtaining a value this large or larger due to
 chance.
tion using the chi squared goodness of fit test (Table 4).  The
null hypothesis of no difference could be accepted at the 0.05
alpha level for 8 of these 12 data sets.  Three of the remaining
four sets had probabilities for acceptance of null that were
greater than 0.02, and only one was less than 0.01 (P<0.001).

Homogeneity of Variances

     Cochran's C statistic was calculated for the same data sets
indicated above (Table 5).  Only 1 of the 12 data sets for
ventilatory rate exceeded the tabulated values at the 0.05 alpha
level.  This indicated that, at that level, the null hypothesis
(that the cells have homogeneous variances) was acceptable for 11
of the 12 intervals tested.

     The same calculations were carried out for the corresponding
amplitude data sets.  It was possible to accept the null hypothesis
at the 0.05 alpha level for 8 of the 12.

Acceptability of ANOVA

     Two-thirds (16) of the 24 data sets tested met both criteria
(normal distribution and homogeneity of variances) for the
acceptability of parametric analyses, and only 3 failed both.  As
this indicates only minor or random departure from normality, the
one-way ANOVA with its relatively robust F test appears to be an
appropriate test for the analysis of these data.
                               38

-------
                 TABLE. 5... . CQCHRAN'S. C STATISTIC.
                                 Interval
                      Pre-
                                  Post-
                                Table
Toxicant.  Data Set. . treatment.  Treatment,  .treatment  Value*  df
Zinc
Zinc

Copper
Copper

MixA
MixA
Rate
Amplitude

Rate
Amplitude

Rate
Amplitude
0.457
0.694

0.429
0.356

0.259
0.514
0.492
0.627

0.468
0.511

0.335
0.588
0.298
0.286

0.565
0.497

0.340
0.506
0.5967  11/3


0.5867  12/3


0.5667  13/3
MixB
MixB
Rate
Amplitude.
0.
0.
477
503. .
0
... o
.389
.429
0
0
.717
.648
0.
5667
13/3

 Values that do not exceed the table value indicate homogeneous
  varianceso
 Largest N-I/Number of variances - 1.


Results of ANOVA

     The results of the one-way analysis of variance tests are
presented in Tables 6 and 7.  In Table 6, the smaller the
probability (P) values are the greater the chance that there is
a detected significant difference among the ventilatory rates or
the average amplitudes with respect to the patterns of applica-
tion of toxicant.

     The Duncan's Multiple Range test was used in an attempt to
better define the indicated differences.  In Table 7, the detected
differences are set off by commas.  However, if one pattern appears
in two or more different groups within a single cell in this
table, these differences are blurred.  The most significant
differences here are for patterns that are set apart from all
other patterns.  For example, during the treatment interval for
the average ventilatory rates for MixB (Table 7), Duncan's
Multiple Range test indicates that the controls are significantly
different from all other patterns and that no difference can be
detected between patterns 1, 2, and 3, a conclusion that can be
confirmed from Figures B-10, B-ll, and B-12.

     It can be seen from the F values in Table 6 that, with one
exception, there is no significant difference indicated between
ventilatory rates for any of the four patterns at the traditonal
alpha level of 0.05 during either the pretreatment or the post-
treatment time intervals.  With only two exceptions,  this holds
                               39

-------
                       TABLE 6.  RESULTS OF ONE-WAY ANOVA.

Interval
Pretreatment
Toxicant
Zinc
Zinc
Copper
Copper
MixA
MixA
MixB
MixB
Data Set
Rate
Amplitude
Rate
Amplitude
Rate
Amplitude
Rate
Amplitude
F
2.77
7.57
1.26
0.50
1.54
1.53
3.58
3.07
P
-------
   TABLE 7.   RESULTS OF DUNCAN'S MULTIPLE RANGE TEST AT THE 95 PERCENT LEVEL
                               OF SIGNIFICANCE.

Interval
Toxicant
Zinc
Zinc
Copper
Copper
MixA
MixA
MixB
MixB
Data Set
Rate
Amplitude
Rate
Amplitude
Rate
Amplitude
Rate
Amplitude
Pretreatment
0 =
0 =
0 =
0 =
0 =
0 =
0 =
0 =
1
1
1
1
1
1
3
1
=3, 0=2*
= 3, 3 = 1, 2
= 2 = 3t
= 2 = 3
= 2 = 3
= 2 = 3
, 3 = 1 = 2
= 2, 3
Treatment
0, 1 =
0 = 3,
0 = 3,
0, 1 =
0 = 3,
0 = 3,
0, 1 =
0, 1 =
2
3
1
2
3
2
2
2
= 3t
= 2, 1
= 2 = 3
= 3
= 2 = 1
= 1
= 3
= 3
Post-treatment
0 =
0 =
0 =
0 =
0 =
0 =
0 =
0 =
3
1
1
2
1
1
1
1
,1 = 2 =
= 2 = 3
,0 = 3 =
- 3, 0 =
= 2 = 3
= 2 = 3
= 2 = 3
= 2v 1 =
3

2
1 = 2



2 = 3

0 = pattern 0 (control); 1 = pattern 1; 2 = pattern 2; 3 = pattern 3.

Significantly different patterns are set off by commas; however, if one pattern
appears in two or more different groups within a single cell, the differences are
less sharp.  For example 0, 1=2=3 indicates that pattern 0 is significantly different
from all other patterns while patterns 1, 2, and 3 cannot be statistically
differentiated.  0=1=2=3 indicates no differences are detectable.

-------
true for the average amplitudes as well.

     It is significant to note that in all instances the pro-
abilities that no differences exist are much smaller (in most
cases an order of magnitude) during the treatment intervals than
during the corresponding pre- and post-treatment intervals.  The
indication being that the significant differences are much
greater during the treatment period, again a conclusion that can
be verified from the graphed data Appendices B and C.

     The results of the Duncan's Multiple Range test (Table 7)
reveal no clear cut pattern during pretreatment and post-treatment
periods.  While differences are indicated for 8 of the 16 data
sets, only 2 of these set a single pattern off from all others,
and these are different patterns in each case.

     During the treatment interval, however, differences are
indicated for all eight data sets with pattern 0 (controls being
differentiated frora: all others most often (four times).  When 0
was not differentiated by itself, it was always (four times)
included in a group with pattern 3,. and. in three of these cases
pattern 3 was also included in other sets.

     It can be concluded that while these results supply no
absolutes, there is a definite trend that indicates statistically
that the effects of sublethal doses of these toxicants are
definitely detectable using, either average ventilatory rates or
average ventilatory amplitudes or preferably both.            -

     The various application patterns chosen for this study cannot
be differentiated as easily.  When toxicant was applied in pattern
1, apparent acclimation occurred, especially during run 4 (Fig.
B-4, C-4).  However, when the toxicant was applied in intermittent
bursts, the reaction to subsequent exposures was often as strong
as it was to the first exposure.  These differences have been
statistically obscured by the extreme smoothing of data' (calcula-
tion of interval means) necessary for these analyses.  Appropriate
methods do exist for the analysis of the ventilatory data for an
individual fish (Box and Tiao, 1975) but are extremely complicated
and time consuming.and are, at present, not practical for quick
analysis of large data sets such as these.
Mu1tip1e Re gr e s s i6n

     The forward stepwise, multiple regression indicates little
or no predictability for any of the five variables used in the
analysis: pattern, standard length, average toxicity ratio, total
dose, or maximum concentration.  In all cases', less than 25 percent
of the variability in ventilatory rate was accounted for by these
independent variables.  None of the one variable models for all
toxicants yielded a significant F ratio.  Furthermore, only


                               42

-------
application pattern in the copper analysis had a significant F
ratio in any of the multiple regression models (two).  However,
this was of no real importance as the overall F ratios for these
two models were not significant.

     It appears, from examination of the figures in Appendices
B and C, that the fish responded more sharply to the higher con-
concentrations of toxicant encountered in patterns 2 and 3 than
to the lower concentrations encountered during pattern 1.  It is
possible that the failure of the statistical analyses to detect
differences due to either application pattern or the maximum
toxicant concentration was due to the extreme smoothing of the
data before it was analyzed.

Possible Causes

     Some workers have indicated that the mechanism by which
heavy metals cause death is related to reduced diffusion capacity
of the gills with resulting lower oxygen levels in the blood
(Skidmore, 1970; Burton et al.,  1972).  Jt would seem reasonable
to hypothesize that such hypoxia might be the cause of the
increased ventilatory rates seen during this study.  Others have
reported increased ventilatory rates with hypoxia (Skidmore,
1964; Gee et al., 1977).  When the water supply was halted to one
of the sensing chambers, the dissolved oxygen level quickly fell
(3 hours) to less than 3.5 mg/fc.  The ventilatory rate increased
from 45 counts per minute to 130-140 counts per minute; however,
the amplitude also increased nearly 300 percent, the opposite of -
the demonstrated amplitude reaction to sublethal levels of heavy
metals.  While this is only preliminary data, it seems to indicate
some other physiological reason for the increased ventilatory
rates.

     The studies indicated above were all carried out with very
concentrated solutions, concentrated enough to cause rapid death,
and emphasized gill damage and hypoxia (Skidmore, 1970; Burton
et al., 1972).  Most chronic studies have emphasized very low
concentrations and have measured parameters such as growth,
reproductive success, and various physiological responses
(Christensen et al., 1972; Brungs et al., 1973; Benoit, 1975).
None of these can presently be correlated with the short term,
sublethal ventilatory effects described here.  Perhaps an investi-
gation of the osmoregulatory changes indicated by Lewis and Lewis
(1971) and Katz (1979) might give a better insight into the
physiological mechanisms involved.
                               43

-------
                           REFERENCES

American Public Health Association (;APHA), American Water Works
     Association, and Water Pollution Control Federation.  1976.
     Standard Methods for the Examination of Water and Wastewater,
     14th Ed. American Public Health Association, New York.

Barnett, J., and D. Toews.  1978.  The Effects of Crude Oil and
     the Dispersant, Oilsperse 43, on respiration and coughing
     rates in Atlantic salmon (Salmo salar) Can. J. Zool.  56:
     307-310.

fielding, D. L.  1929.  The Respiratory Movements of Fish as an
     Indicator of a Toxic Environment.  Trans. Am. Fish. Soc.
     59:238-245.

Benoit, D. A.  1975.  Chronic Effects of Copper on Survival,
     Growth, and Reproduction of the Bluegill (Lepomis macrochirus)
     Trans. Am. Fish. Soc.  104:353-358.

Box, G. E. P., and G. C. Tiao.  1975.  intervention Analysis with
     Applications to Economic and- Envirpnniental- Problems.  J. Am.
     Stat. Assoc.  70:70-79".

Brungs, W. A., E. N. Leonard, and J. M. McKim.  1973.  Acute and
     Long-Term Accumulation .of Copper by the Brown Bullhead,
     Ictalurus riebulosus.  J. Fish. Res. Board Can.  30:583-586.

Burton, D. T., A. HI Jones)'and'J. Cairns, Jrv  1972. ..Acute Zinc
     Toxicity to Rainbow Trout (Salmo gairdneri): Confirmation
     of the Hypothesis that Death.is Related to Tissue Hypoxia.
     J. Fish. Res. Board Can.  29': 1463-1466.

Cairns, J., Jr., K. L. Dickson, R. E. Sparks, and W."T. Waller.
     1970.  A Preliminary Report on Rapid Biological Information
     Systems for Water Pollution Control.  J. Water Pollut.
     Control Fed.  42:685-703.

Cairns, J., Jr., D. Gruber, K. L. Dickson, A. C. Hendricks
     and W. H. van der Schalie.  1977.  Developing an On-Site
     Continuous Biological Monitoring System for the Chemical
     Industry.  Pages 285-294 in L. Delpino and A. Krigman,
     Eds. ,. Proceedings of Fifth Annual Industrial Pollution
     Conference, Water and Wastewater Equipment Manufacturers
     Assoc.,  Inc., McLean, Virginia.

Cairns, J., Jr., and A. Scheier.  1957a.  The Effects of Tempera-
     ture  and Hardness of Water upon the Toxicity of Zinc to the
     Common Bluegill (Lepomis macrochirus Raf.). Notulae
     Naturae  299.
                                44

-------
Cairns, J., Jr., and A. Scheier.  1957b.  The Effects of Periodic
     Low Oxygen upon the Toxicity of Various Chemicals to Aquatic
     Organisms.  Proceedings of the 12th Industrial Waste Confer-
     ence, Purdue Univ. Eng. Bull.  94:165-176.

Cairns, J., Jr., and K. W. Thompson.  In press.  A Computer Inter-
     faced Toxicity Testing System for Simulating Variable
     Effluent Loading.  Proceedings of the Second Symposium on
     Process Measurements for Environmental Assessment.

Cairns, J., Jr., K. W. Thompson, J. D. Landers, Jr., M.  T. McKee,
     and A. C. Hendricks.  1980.  Suitability of Some Freshwater
     and Marine Fishes for Use with a Minicomputer Interfaced
     Biological Monitoring System.  Water Resour. Bull.   16:
     421-427.

Cairns, J., Jr., and W. H. van der Schalie.  In press.  Biological
     Monitoring, Part I:  Early Warning Systems.  Water Research.

Camougis, G.  1960.  Recording Bioelectrical Potential from
     Aquatic Animals.  Turtox News 38:156-157.

Carlson, R. W., and R. A. Drummond.  1978.  Fish Cough Response -
     a Method for Evaluating Quality of Treated Complex Effluents.
     Water Res. 12:1-6.

Chizar, D., M. Moody, and J. T. Windell.  1972.  Failure of
     Bluegill Sunfish, Lepomis macrochirus. to Habituate to
     Handling.  J. Fish. Res. Board Can.  29:576-578.

Christensen, G. M., J. M. McKee, W. A. Brungs, and E. P. Hunt.
     1972.  Changes in the Blood of the Brown Bullhead (Ictalurus
     nebulosus (Lesueur)) Following Short and Long Term Exposure
     to Copper (II).  Toxicol. Appl. Pharmacol.  23:417-427.

Davis, J. C.  1972.  Sublethal Effects of Bleached Zraft Pulp
     Mill Effluents on Respiration and Circulation in Sockeye
     Salmon.  J. Fish. Res. Board Can.  30:369-377.

Drummond, R. A., and R. W. Carlson.  1977.  Procedures for
     Measuring Cough (gill purge) Rates of Fish.  Ecol.  Res.
     Ser.  EPA-600/3-77-133.

Drummond, R. A., G. F. Olson, and A. R. Batterman.  1974.  Cough
     Response and Uptake of Mercury by Brook Trout, Salvelinus
     fontinalis, Exposed to Mercuric Compounds at Different
     Hydrogen-Ion Concentrations.  Trans. Am. Fish. Soc.  103:
     244-249.                                       .

Drummond, R. A., W. A. Spoor, and G. F. Olson.  1973.  Some
     Short-Term Indicators of Sublethal Effects of Copper on


                               45

-------
     Brook Trout, Salvelinus fontinalis.  J. Fish. Res. Board
     Can.  30:698-701.

Finney, D. J.  1971.  Statistical Methods in Biological Assay,
     2nd Ed. Griffin Press, London, England.

Fee, J. H., R. F. Tillman, and H. J. Smart.  1977.  Reactions
     of Some Great Plains Fishes to Progressive Hypoxia.  Can.
     J. Zool.  56:1962-1966.'

Gruber, D., J. Cairns, J; , K. L. Dickson, and A. C. Hendricks.
     1979.  A Cinematographic Investigation into the Fish's
     Bioelectric Breathing Signal.  J. Fish. Biol.  14:429-436.

Gruber, D., J. Cairns, Jr., K. L. Dickson, R. Hummel, III,
     A. Maciorowski, and W. H. van. der. Schalie.  1977.  An
     Inexpensive, Noise-Immune Amplifier Designed for Computer
     Monitoring of Ventilatory Movements of Fish arid Other
     Biological Events.  Trans. Am. Fish. Soc.  106:497-499.

Hendricks, A-.--C. ,.X. L. Di-cksony and J. Cairns, Jr.  1:977.  A
     Mobile Fish Bioassay 'Unit for Effluent Testing.  Report
     to Manufacturing Chemists Association.

Hughes, G. M.  1975.  Coughing in the Rainbow Trout (Salmo
     gairdneri) and the Influence of Pollutants.  Revue Suisse
     Zool.  82:47-64.

Katz, B.   1979.  Relationship of the Physiology of Aquatic
     Organisms to the Lethality of Toxicants:  A Broad Overview
     with  Emphasis on Membrane  Permeability.  Pages 62-76 in
     L. L. Marking and R. A. Kimerle, Eds.  Aquatic Toxicology,
     STP 607, American Society fpr testing and Materials,
     Philadelphia, Pa.

Kleerekoper, H., and K. Sibakin.  1956.  An Investigation of
     the Electrical "Spike" Potentials Produced by the Sea
     Lampregy (Petromyzon marinus) in the Water Surrounding
  ••• the Head Region.  J.''Fish. Res'. Board Can.  13:375-383.

Lewis,; S.  D., and W. M. Lewis.  1971.  The Effect of Zinc and
     Copper on the Osmolality of Blood Serum of the Channel
     Catfish, Totalurns' puhctatus Rafinesque, and Golden Shiner,
     Notemigonus crysbleucas Mitchill.  Trans. Am. Fish. Soc.
     100:639-643.

Lunn, C. R., D. P. Toews, and D. J. Pree.  1975;  Effects of
     Three Pesticides on Respiration, Coughing, and Heart
     Rates of Rainbow Trout (Salmo gairdneri Richardson).
     Can.  J. Zool.  54:214-219.
                               46

-------
Maki, A. W.  1979.  Respiratory Activity of Fish as a Predictor
     of Chronic Fish Toxicity Values for Surfactants.  Pages
     77-95 in L. L. Marking and R. A. Kimerle,  Eds., Aquatic
     Toxicology, STP 667, American Society for Testing and
     Materials, Philadelphia, Pa.

Marking, L. L.  1977.  Method for Assessing Additive Toxicity
     of Chemical Mixtures.  Pages 99-108 iji F. L. Mayer and
     J. L. Hamelink, Eds., Aquatic Toxicology and Hazard
     Evaluation, STP 634, American Society for Testing and
     Materials, Philadelphia, Pa.  307 pp.

Meuwis, A. L., and M. J. Heuts.  1957.  Temperature Dependence
     of Breathing Rate in Carp.  Biol. Bull., Woods Hole
     112:97-107.

Morgan, E. L., K. W. Eagleson, R. Herrmann, and N. D. McCollough,
     Jr., 1979.  Automated Biomonitoring Applications to Remote
     Water Quality Stations and Satellite Data Retrieval:  New
     Developments in Acheiving Real-Time Biosensing for Watershed
     Management.  Ninth Intersociety Conference in Environmental
     Systems.

Morgan, W. S. G.  1977.  An Electronic System to Monitor the
     Effects of Changes in Water Quality on Fish Opercular Rhythms.
     Pages 38-55 in J. Cairns, Jr., K. L. Dickson and G. F.
     Westlake, Eds., Biological Monitoring of Water and Effluent
     Quality, STP 607, American Society for Testing and Materials,
     Philadelphia, Pa.

Morgan, W. S. G., and P. C. Kuhn.  1974.  A Method to Monitor
     the Effects of Toxicants upon Breathing Rate of Largemouth
     Bass (Micropterus salmoides Lacepede).  Water Res. 8:67-77.

Nunn, G. L., J. Cairns, Jr.,  and D. Gruber.  In press.  The
     Distribution of Ventilatory Frequencies Over Time of the
     Bluegill, Lepomis macrochirus.  J. Interdiscipl. Cycle Res.

Pasztor, V. M., and H. Kleerekoper.  1962.  The Role of the
     Gill Filament Musculature in Teleosts.  Can. J. Zool.
     40:785-806.

Patrick, R., J. Cairns, Jr.,  and A. Scheier.  1968.  The
     Relative Sensitivity of Diatoms, Snails and Fish to
     Twenty Common Constituents of Industrial Wastes.  Prog.
     Fish-Cult. 31:137-140.

Peltier, W.  1978.  Methods for Measuring the Acute Toxicity
     of Effluents to Aquatic Organisms.  EPA-600/4-78-012,
     U. S. Environmental Protection Agency.
                               47

-------
Pickering, Q. H.,  and C. Henderson.  1966.  The Acute Toxicity
     of Some Heavy Metals to Different Species of Warmwater
     Fishes.  Air Water Pollut. Int. J. 10:435-463.

Roberts, J. L.  1964.  Metabolic Responses of Freshwater Sunfish
     to Seasonal Photoperiods and Temperatures.  Helgoloender
     Wissenschafteiche Meeresuntersuchunger 9:459-473.

SAS Institute, Inc.  1979.  SAS User's Guide.  SAS Institute,
     Raleigh, N.C.

Saunders, R. L.  1961.  The Irrigation of the Gills in Fishes,
     I.  Studies of the Mechanism of Bronchial Irrigation.
     Can. J. Zool.  39:637-653.

Schaumburg, F. D., T. E. Howard, and C. C. Walden.  1967.  A
     Method to Evaluate the Effects of Water Pollutants on
     Fish Respiration.  Water Res.  1:731-737.

Sellers, C. M., A. G. Heath, and M. L. Bass.  1975.  The Effect
     of Sublethal Concentrations of Copper and Zinc on
     Ventilatory Activity, Blood Oxygen and pH in Rainbow Trout
     (Salmo gairdneri).  Water Res.  9:401-408.

Siegle, S.  1956.   Nonparametric Statistics for the Behavioral
     Sciences.  McGraw-Hill, New York.  312 pp.

Skidmore, J. F.  1964.  Toxicity of Zinc Compounds to Aquatic..
     Animals, with Special Reference to Fish.  Quart. Rev. Biol.
     39:227-248.

Skidmore, J. F.  1972.  Respiration and Osmoregulation in Rainbow
     Trout with gills damaged by Zinc Sulphate.  J. Exp. Biol.
     52:484-494.

Sorensen, E. M. B., and R. D. Burkett.  1978.  Respiratory
     Responses to Instant Thermal Stress Under Oxygen- Enriched
     Conditions in Centrarchids.  Texas J. Sci.  30:337-345.

Sparks, R. E., J.  Cairns, Jr., R. A. McNabb, and G. Suter, II.
     1972.  Monitoring Zinc Concentrations in Water Using the
     Respiratory Response of Bluegills (Lepomis macrochirus
     Rafinesque).   Hydrobiologia.  40:361-369.

Spoor, W. A., T. W. Neiheisel, and R. A. Drummond.  1971.  An
     Electrode Chamber for Recording Respiratory and Other
     Movements of Free-Swimming Animals.  Trans. Am. Fish. Soc.
     100:22-28.
                              48

-------
Sprague, J. B.  1973.  The ABC's of Pollutant Bioassay Using Fish.
     Pages 6-30 in J. Cairns, Jr., and K. L. Dickson, Eds.,
     Biological Methods for the Assessment of Water Quality, STP
     528, American Society for Testing and Materials, Philadelphia,
     Pa.  256 pp.

Thomas, R. E., and S. D. Rice.  1975.  Increased Opercular Rates
     of Pink Salmon (Oncorhynchus  gorluscha) Fry after Exposure
     to the Water Soluble Fraction of Prudhoe Bay Crude Oil.  J.
     Fish. Res. Board Can.  32:2221-2224.

Thomas, R. E., and S. D. Rice.  1979.  The Effect of Exposure
     Temperature on Oxygen Consumption and Opercular Breathing
     Rates of Pink Salmon Fry Exposed to Toluene, Naphthalene, and
     Water-Soluble Fractions of Cook Inlet Crude Oil and No. 2 Fuel
     Oil.  Pages 39-52 in W. B. Vernberg, F. P. Thurberg, A.
     Calabrese, and F. J. Vernberg, Eds., Functional Responses.
     Academic Press.

Thompson, K. W.,  K. L. Dickson, D. Hooley, J. Cairns, Jr., and
     W. H. van der Schalie.  1978.  A Computer Monitored Aquatic
     Bioassay System.  Pages 61-66 iji J. B. Carberry, Ed.,
     Proceedings of Tenth Mid-Atlantic Industrial Waste Conference,
     University of Delaware.

Trama, F. B.  1957.  The Acute Toxicity of Copper to the Common
     Bluegill (Lepomis macrochirus Raginesque).  Notulae Naturae 257.

Trama, F. B., and R. J. Benoit.  1960.  Toxicity of Hexavalent
     Chromium to Bluegills.  J. Water Pollut. Control Fed.
     32(8):868-877.

Tsai, C. and J. A. McKee.  1980.  Acute Toxicity to Goldfish of
     Mixtures of Chloramines, Copper, and Linear Alkylate
     Sulfonate.  Trans. Am. Fish. Soc.  109:132-141.

van der Schalie,  W. H.  1977.  The Utilization of Aquatic
     Organisms for Continuous and Automatic Monitoring of the
     Toxicity of Industrial Waste Effluents.  Ph.D. Dissertation.
     Virginia Polytechnic Institute and State University,
     Blacksburg,  Va.

van der Schalie,  W. H., 1980.  A New Technique for Automatic
     Monitoring of Fish Ventilatory Patterns and Its' Possible
     Use in Screening Tests for Chlorine Toxicity.  Pages 233-242
     in J. G. Eaton, P. R. Parrish, and A. C. Hendricks, Eds.,
     Aquatic Toxicology, STP 707, American Society for Testing
     and Materials, Philadelphia, Pa.

van der Schalie,  W. H., K. L. Dickson, G. F. Westlake, and
     J. Cairns, Jr.  1979.  Fish Bioassay Monitoring of Waste
     Effluents.  Environ. Manage.  3:217-235.

                               49

-------
Westlake, G. F.,  and W. H. van der Schalie.  1977.  Evaluation
     of an Automatic Biological Monitoring System at an
     Industrial Site.  Pages 30-37 in J. Cairns, Jr., K. L.
     Dickson, and G. F. Westlake, Eds.,  Biological Monitoring
     of Water and Effluent Quality, STP 607, American Society
     for Testing and Materials, Philadelphia, Pa.
Winer, B. J.
     Design.
1971.  Statistical Principles in Experimental
McGraw Hill, New York.
Young, Y.  1971.  EMG activity in Tench (Tinea tinea, L.)
     Gill Lamellae and its Association with Coughing.  J.
     Phys. 215:37P-38P.
                               50

-------
                           APPENDIX A

     RESULTS OF WATER ANALYSES AND INDIVIDUAL TOXICITY TESTS
TABLE A-l.  RESULTS OF CHEMICAL ANALYSES GIVEN AS THE AVERAGE
            OF THE FIVE ANALYSES FOR EACH FISH BIOASSAY CHAMBER
Test
number
1
(Zinc) '




2
(Zinc)




3
(Copper)




4
(Copper)




5
(Nickel)




Tank Conductivity Alkalinity
number micromhos/cm pH (mg/1)
1
2V
3IV
4III
511
I
1
2V
3IV
4III
511
I
1
2V
3IV
4III
511
I
1
2V
3IV
4III
511
I
1
2V
3IV
4III
511
I
110
123
131
150
179
204
109
115
118
121
129
139
106
106
106
106
106
106
110
111
112
113
115
117
118
128
134
147
162
194
7.5
7.4
7.4
7.3
7.3
7.2
7.2
7.2
7.1
7.0
7.0
6.9
7.1
7.1
7.1
7.0
6.9
6.9
7.1
7.1
7.1
7.1
6.9
6.9
7.1
7.1
7.1
7.1
7.1
7.1
26.8
28.0
28.4
29.2
29.2
29.2
31.6
32.0
31.2
30.0'
30.4
30.4
23.2
24.0
24.0
24.0
24.0
24.0
28.0
31.2
30.4
30.4
30.8
30.8
30.0
32.8
33.2
33.2
32.8
32.8
Total
hardness
(mg/1 CaC03)
22.4
28.8
32.8
39.6
48.0
59.2
23.6
27.2
28.4
31.2
35.6
45.6
28.8
. 28.8
28.8
28.4
28.0
28.8
22.0
21.2
23.3
23.6
24.4
24.8
36.6
34.6
39.0
32.2
31.8
33.2
                               51

-------
TABLE A-l (cont.)
Test
number
6
(Nickel)




7
(Chromium)




8
(Chromium)




9
(MixA)




10
(MixA)




11
(MixB)




Tank Conductivity
number micromhos/cm pH
1 .
2V
3IV
4III
511
I
1
2V
31 V
4III
511
I
1
2V
3IV
41 II
511
I
1
2V
3IV
: 41 IT "
511
I
1
" 2V
' 3IV
4III .
511- :
I
1
2V
3IV
4III
511
I
125
149
163
189
225
320
117
565
876
1496
2052
.3352
111
256
385
628
946
1198
111
114
114
117 .
120
126
118
123
125
.... "127
130
136
131
263
332
506
732
1106
7,1.
7.1
7.1
7.1
7.1
7.0
7.1
7.6
7.8
7.9
8.2.
..8.3.
..7.1
7.2
7.4
7.5
7.6
7.8
7.0
7.0
7.0
7.0
6.9
'•*' 6.8
••" 7.1
7.1
7.1
'-....• ••-•7.1"
7.0
6.9
6.8
7.1
7.2
7.4
7.5
7.7
Alkalinity
(mg/1)
36.0
35.2
35.6
36.0
35.8
35. 8
30.4
97.2
139.6
221.2
361.6
551.0
40.8
63.6
83.8
110.8
180.4
.268.0
30.4
30.0
30.0
30.0
30.0
" 30.0
•--•- 30.4
30.8
31.6
32.8
32.4 -.
32.8
31.6
53.8
65.0
94.8
131.6
218.0
Total
hardness
(mg/1 CaC03)
71.4
76.0
65.2
67.6
68.4
70.8
22.0
19.6
22.4
20.8
20.8
20.8
44 . 4
43.6
42.4
42.8
42.4
42.0
21.6
21.2
21.4
22.2
24.6
.. 26-6
29.2
32.4
32.8
32.8
34 . 8
34.8
44.8
40.4
38.8
41.6
43.6
44.6
                               52

-------
TABLE A-l (cont.)
                                                       Total
   Test      Tank   Conductivity       Alkalinity    hardness
  number    number  micromhos/cm   pH    (mg/1)     (mg/1 CaCO.,)
                                                             «J


  12         1           119      7.1     33.0         35.0
(MixB)       2V          310      7.3     57.2         43.0
             3IV         421      7.4     75.2         41.0
             4III        668      7.5    114.0         37.6
             511        1106      7.6    171.6         38.6
              I         1719      7.6    269.2         35.6
                               53

-------
TABLE A-2.  TOXICANT CONCENTRATIONS AND  SURVIVAL DURING TOXICITY
            TESTS  ....••      .       "             .


Test
number
1
(Zinc)




2
(Zinc)




3
(Copper)




4
(Copper)




5
(Nickel)




6
(Nickel)






Tank
number
1 (Control)
2V
3IV
4III
511
I
1 (Control)
2V
3IV
4III
511
I
1 (Control)
2V
3IV
4III
511
•-, I . .
1 (Control)
2V
3IV
4III
511
I
1 (Control)
2V
3IV
41 1 1
5LI- -- -
" I • 	
1 (Control)
2V
3IV
4III
511
I

Average
Zn++
0.0
2.6
3.6
6.2
10.9
23.2
0.1
1.2
2.1
3.5
5.7
9v9 ••
_ ..•_..
••'• -..,. _ ..
—
—
. -
-• *-.-->•".'"'.• .
_
- '
-
-
-
—
_
. - •
-
-
—
-
_
-
—
_
—
-

concentration (mg/1)
++ ++ +6
Cu Ni Cr
_ — • «
_ . _ _
_ _ _
_ _ _
_ _. _
_ _ _
_ ' _ _
.•*•'.— -
_ _ _
_ _ _
_ • _ _
".. - • _ : _
• o.o : -
0.3 , - -
0.5
•0.8..-
1.4
V-2.4,..,,., ,.,.,, ........ .f-,,.,.
o.o ' "•••'•--. _ • ': :
0.5 - -
0.7 - -
1.3
2.3 -
4.0 •;..' - ....'•"' \ -
0.0
2.3
3 . 3
7.7
-12.0
20.9
0.1
7.1
10.0
20 . 3
28.4
50.0 -

# Surviving
at 96 hr
(of 10)
10 '
9
5
0
0
0
10
7
7
8
1
0
10
10
10
10
0
• - . ,0,..
10
9
7
3
0
o
10
10
10
10
9
6
10
10
10
4
3
0
                                54

-------
TABLE A-2 (cont.)
Average concentration
Test
number
7
(Chromium)




8
(Chromium)




9
(MixA)




•? '.-lo'1:.
(MixA)




11 •",.
(MixB) '




12
(MixB)




Tank
number
1 (Control)
2V
3IV
4III
511
I
1 (Control)
2V
3IV
4III
511
I
1 (Control)
2V
3IV
4III
511
I
1 (Control)
2V
3IV
4III
511
I
1 (Control)
2V
,3IV
4III
511
I
1 (Control)
2V
3IV
4III
511
I
Zn++
_
-
-
-
—
-
_
-
-
-
-
-
0.0
0.4
0.6
1.2
1.8
3.1
0.0
0.4
0.6
1.2
2.0
3.7
0.0
0.4
0.6
1.0
1.9
3.0
0.0
0.4
0.7
1.3
2.0
3.1
Cu++
_
-
-
-
-
-
_
-
-
-
-
-
0.0
0.1
0.2
0.3
0.5
0.9
0.0
0.1
0.2
0.4
0.5
1.0
0.0
0.1
0.2
0.5
0.8
1.4
0.0
0.3
0.5
0.8
1.6
2.6
Ni++
_
-
-
-
—
-
_
-
-
-
-
-
_
-
-
_
-
- •-
; '• _
-
' - •
-
-
-
0.0
3.0
4.8
8.9
14.6
24.3
0.0
5.9
10.3
16.9
31.6
46.7
(mg/1)
Cr+6
0.1
37.6
62.8
119.8
181.0
289.0
0.1
66.8
109.6
199.8
331.2
525.2 ...
_
-
-
—
-
- •
• '_ / ;"'
. :_
-
- '
-
- ' '•
0.0
29.3
36.6
72.4
99.6
181.8
0.0
30.2 ,
49.9
78.1
135.1
226.0
# Surviving
at 96 hr
(of 10)
8*
9*
10
8
3
0
10
.. -.8
8
.0 . . ...
0
• .,..- . 	 	
10
10
10
5
0
0
v . VLO •••
': -lo - '
'9
-6t
8
0
10
8
10
7
0
0
8*
9
9
1
0
0
*Two, one, and two disregarded due to behavioral death
tOnly nine fish at the start, counting error
                               55

-------
                           APPENDIX B



     Graphic representation of the average ventilatory ra-tes



for all 12 ventilatory tests.  See text Figures 12, 13, and 14



and text table 1 for further details.
                               56

-------
01,
                0  24  48  72
                          86  120 144 168  192 216

                         TIME (HOURS)
Figure B-l   Average ventilatory rates  for ventilatory test 1.  Toxicant = Zn
pattern  =  1,  concentration = 2.5 mg/1.           .'..;.
                                                                                      ++

-------
Oi !
ooi
                                                               k*K^CJC?y*Fi
                          Tl.  M 120 .' 144

                           TIME (HOURS)
    Figure B-2.  Average  ventilatory rates for ventilatory  test  2.   Toxicant = Zn   ,
    pattern = 2, concentrations = 2.5,  5.7, 8.1 mg/1.

-------
 01 I
 CD j
,  . I
                    72 M

                     TIME (HOURS)
     Figure B-3,   Average ventilatory rates  for  ventilatory test. 3.  Toxicant  =  Zn

     pattern  =  3,  concentrations = 2.9, 5.7,  7.7 mg/1.

-------
                     46  72  96  120 144  166 192  216

                         TIME (HOURS)
Figure B-4.   Average ventilatory rates for ventilatory test 4.   toxicant = Cu++,
pattern =  1,  concentration =  0.5 mg/l.

-------
Ci;
M,
^(*>^i^j<^^
                                          >O£ZL' '  '^6
                                           ^?"~^~—7~/ • HSH.i.
                        72 ' 96  120 M«

                         TIME (HOURS)
    Figure B-5.  Average ventilatory rates for ventilatory test 5.  Toxicant  =  Cu++>
    pattern = 2, concentrations  =  0.5,  1.0,  1.5 rng/1.  Representative strip chart  records
    for number 5, 6, 7, and  8  can  be found in Appendix D.      .-: '

-------
CO
                                  itellL^Xl
                      72  96  120 . 144  ISO , 192

                       TIME (HOURS)
    Figure B-6.  Average ventilatory rates for ventilatory test 6.  Toxicant  = Cu++,
    pattern = 3, concentrations = 0.5, 1.1, 1.3 tng/1.  *Fish 6, a control,  received  a
    partial dose of toxicant due to splash and was not included in the statistical analysis,

-------
GO I
                   2A 46  72
                          96  l» 144

                         TIME (HOURS)
                                  KB 192  36
    Figure B-7.   Average ventilatory  rates for.ventilatory test 7
    pattern = 1,  concentrations =0.6 mg/1 Zh  ; 0.2 mg/1  Cu  .
Toxicant =  MixA,

-------
cn
                                                 -^-V^^^
                    M  43  72
                           W 120  144 oa  192 2K
                          TIME(HOURS)
     Figure B-8.   Average ventilatory rates for ventilatory test  8    Toxicant - MixA,
     pattern  =  2,  concentrations = 0.6,  1.3,  1.8 mg/1 Zn++; 0.2,  0.4,  0.6 mg/1 Cu   .

-------
O) '
01,
                                     •%W'Wty&7 '   A«
                                     X&^^         E
                  O  SO 40  72 96  120 144  168 192

                          TIME (HOURS)
    Figure  B-9.   Average ventilatory rates for ventilatory test 9.  Toxicant = MixA,
    pattern = 3,  concentrations = 0.6, 1.4, 1.8 mg/1 Zn++; 0.2, 0.5,  0.6 mg/1 Cu++.

-------
05]
01'
                                ISO .,192 218
                        TIME (HOURS)
Figure B-10.  Average ventilatory rates for ventilatory  test 10.   Toxicant
pattern = 1, concentrations = 0.3 mg/1 Zn++; 0.1 mg/1  Cu++;  2.5 mg/1 Ni**;
25.5 mg/1 Cr+6.
                                                                                 = MixB,

-------
                                             7^r^>
                    72  W I20  144 08  192
                     TIME (HOURS)
Figure B-ll.   Average ventilatory rates  for  ventilatory test 11.  Toxicant - MixB,
pattern = 2,  concentrations = 0.4,  0.8,  1.1  mg/1 Zn++; 0.1, 0.2, 0.2+ mg/1 Cu++;
3.3, 5.1, 8.0 mg/1 Ni++;  22.5,  44.9,  58.7  mg/1  Cr+6.

-------
                     72 96  120 144  ISO 192

                      TIME(HOURS)
Figure B-12.  Average  ventilatory rates for ventilatory  test 12.   Toxicant = MixB,
pattern = 3, concentrations = 0.4, 0.8, 1.1 mg/1 Zn++; 0.1,  0.2,  0.2+ mg/1 Cu++;
2.3, 5.3, 6.9 mg/1  Ni++;  22.3,  60.6, 77.0 mg/1 CrT6.

-------
                          APPENDIX C



     Graphic representation of the average signal amplitude



for all 12 ventilatory tests.   See text Figures 12, 13,  and



14 and text table 1 for further details.
                               69

-------
•v)
o;
                          46  72  96  120 144

                             TIME (HOURS)
    Figure C-l.  Average  ventilatory signal  amplitudes for ventilatory test 1,
    Toxicant = Zn++, pattern =1.

-------
                       TIME (HOURS)
Figure C-2.  Average  ventilatory signal amplitudes for ventilatpry test 2.
Toxicant = Zn++, pattern = 2.             .               :   .''.'••

-------
-J
to
                                      '/^mzw    A*, =./
                                      i^^^xov^^x /  v.
                          86  120 144

                         TIME (HOURS)
   Figure C-3.   Average ventilatory signal amplitudes for ventilatory test 3.
   Toxicant  =  Zn++,  pattern =3.

-------
CO!
                                 120  144

                             TIME (HOURS)
   Figure C-4    Average ventilatory  signal amplitudes for ventilatory test 4
   Toxicant =  Cu++,  pattern =1.                            •

-------
                       S& . 120 144
                     ; TIME (HOURS)
Figure C-5.   Average ventilatory signal  amplitudes for ventilatory test 5,
Toxicant = Cu   ,  pattern = 2.

-------
en!
                       2K>

                       £00
                       ISO
                         ^^^_^^^
                 S^&J^J^^^
                        2  86 120  144 168

                        TIME (HOURS)
   Figure  C-6.   Average ventilatory signal amplitudes  for ventilatory  test  6.
   Toxicant  = Cu++,  pattern = 3.  *Fish 6, a control,  received.a  partial dose  of
   toxicant  due to water splash and was not included in  the' statistical analysis,

-------
                   .46 .-72  9S  IEO 144  KS . 192  2t6

                   I'  : TIME (HOURS)
Figure C-7.   Average ventilatory signal  amplitudes for  ventilatory test  7.
Toxicant  =  MixA, pattern  =  1.

-------
                    TIME (HOURS)
Figure C-8.   Average ventilatory  signal amplitudes for v>entilatory test 8.
Toxicant = MixA,  pattern = 2.                      ..';  /  | '-. .:;    .

-------
2OO
ISO
                                                 ^^
00 ,
                        72 . 86
                         TIME (HOURS)
    Figure C-9.   Average ventilatory  signal amplitudes for  ventilatory test 9.
    Toxicant = MixA,  pattern = 3.

-------
                      96 120  144

                    TIME (HOURS)
Figure C-10.   Average ventilatory signal  amplitudes for ventilatory  test 10,
Toxicant = MixB,  pattern =1.

-------
CO,
O!
                       TIME (HOURS)
    Figure C-ll.  Average ventilatory signal amplitudes for ventilatory test 11.
    Toxicant = MixB, pattern  =  2.

-------
Hi
                        TIME (HOURS)
    Figure C-12.   Average veiitilatory signal  amplitudes for ventilatory  test  12.
    Toxicant =  MixB,  pattern =? 3.

-------
                          APPENDIX D



     Examples of stressed and unstressed inter- and intra-



individual variation as illustrated by strip chart recordings,



These figures can be coordinated with the graphs of average



ventilatory rate in Figure B-5 and the average signal ampli-



tude in Figure C-5.
                               82

-------
               0-
               -3
                o-
              O
              >
              I- '
              <
                o-
               -3J
                0 -
               -3J
                            10         20
                               SECONDS
                                                30
Figure D-l.  Examples of the appearance of the ventilatory signals
as they appear on a strip chart recorder.  These recordings are
for fish 6 (control) from run 5.  The voltages indicated are
arbitrary since the strip chart recorder further amplifies the
signal that is being sent to the computer.  The average rates and
amplitudes recorded by the computer during the 15-min period from
which these recordings were made and the total hours elapsed since
the beginning of interval 2 are:  A.  14 cts/min, 3.6 volts, 22.5
hr; B.  27, 2.8, 44.5; C.  49, 3.1, 74.5; D.  25, 5.3, 94.5.
                               83

-------
              0-
              -3J
              0-
            UJ
            IT
              0-
              -3J

              »3
              0-
              -3J
10         20
  SECONDS
                                                30
Figure D-l (cont.).  Ventilatory signals  for  fish  6  (control)  from
run 5.  The computer values for the average rates  and amplitudes
and the elapsed time are:  E.  35 cts/min, 6.7  volts,  94.5 hr;
F.  25, 5.9, 97; G.  21, 5.8, 115.5; H.   31,  4.7,  123.
                               84

-------
              *5 n
              0-
              -5

              «5
             O
             
-------
 -3

 •2 -i



  0 •


en -2


o
                 J«ll^^
               0-



              -2-

              *3 -





               0 -
                            10         20
                              SECONDS
                                    30
Figure D-2 (cont.).  Ventilatory signals for fish 5  (treated)  from
run 5.  The computer values for the average rates and amplitudes
and the elapsed time are:  D.  25 cts/min, 4.4 volts, 94.5 hr;
E.  56, 1.8, 97 (after 1 hr of toxicant exposure, concentration =
0.6 mg/1 Cu++); F.  127, 3.8, 107.5 (after 11.5 hr of exposure,
1.9 mg/1); G.  32, 3.4, 115.5 (after 19.5 hr exposure, 0.6 mg/1);
H.  42, 6.1, 123 (between toxicant applications after 8 hr of
recovery time).
                               86

-------
               *3 1
               0-
               -3

               »3 1
               0 -
              UJ
              en
               -3
               o-
                            10         20
                              SECONDS
                                                 A
                                                30
Figure D-3.  Ventilatory signals  for  fish 7  (treated)  from run 5.
The computer values for the average rates and amplitudes and the
elapsed time are:  A.  18 cts/min, 6.1  volts,  22.5 hr;  B.   18,
6.0, 44.5; C.  21, 6.3, 74.5; D.   20, 5.3, 94.5.
                               87

-------
             + 3 n
              0-
             '-3J

             + 3 -i
	|UU	\J\tAs-Wfs**^fff^	k/v	ryy*	^
                 A>^W\fV^Nv-^^
            UJ
            cc
              0-
             -3

             + 3 i
              0 •
             -3 J
                            10
                                      . 2 Q
                                                 30
                              SECONDS
Figure D-3 (cont.).  Ventilatory signals for fish 7  (treated)  from
run 5.  The computer values for the average rates and amplitudes
and the elapsed time are:   E.   46 cts/min, 2.1 volts, 97 hr  (after
1 hr of toxicant  exposure,  concnetration = 0.5 mg/1  Cu"1"1"); F.   75,
2.2, 107.5 (after 11.5  hr  of exposure, 1.4 mg/1); G.  13,  1.7,
115.5 (after 19.5 hr exposure,  0.5 mg/1); H.  36, 6.7,  123 (between
applications after 8 hr of recovery time).
                                88

-------
             +5 n
              -5

            OT + 3
            O
            >
              -5
                           10
                                     20
                                               30
                              SECONDS
Figure D-4.  Ventilatory signals for fish 8 (treated) from run  5,
The computer values for the average rates and amplitudes and the
elapsed time are:  A,  37 cts/min, 7,7 volts, 22.5 hr; B.  40,
5.8, 44.5; C.  68, 13.9, 74.5.
                               89

-------
                           10
                              SECONDS
                                                30
Figure D-r4 (cont.).  Ventilatory signals for fish 8 (treated) from
run 5.  The computer values for the average rates and amplitudes
and the elapsed time are:  D.  38 cts/min, 7.7 volts, 94«5 hr;
E.  57, 3,6, 97 (after 1 hr of toxicant exposure, concentration =
0.5 mg/1 Cu++); F.  20, 1.8, 107.5 (after 11.5 hr of exposure,
1.3 mg/1); G.  26, 1.8, 115.5 (after 19.5 hr exposure, 0.5 mg/1);
H.  44, 6.9, 123 (between toxicant applications after 8 hr of
recovery time).
                               90

-------

-------