the use of fish as sensors in industrial waste lines to prevent fish kills
TRANSCRIPT
Hydrobiologia, vol. 41, 2, pag . 151-167, 1973
The Use of Fish as Sensors inIndustrial Waste Lines to Prevent Fish Kills
by
JOHN CAIRNS, JR ., RICHARD E. SPARKS & WILLIAM T. WALLER
Center for Environmental Studies and Department of Biology,Virginia Polytechnic Institute and State University
Blacksburg, Virginia 24061
The report of the Council on Environmental Quality (1970)repeatedly stresses the need for the development of predictive,simulative, and managerial capabilities to combat air and waterpollution. The last capability depends on the first two . For example,the effects of every waste put into a river will have to be predicted ifthe river is to be managed as a system wherein industrial use of thewater does not preclude other uses such as recreation and municipalwater supply . In addition, the effects of alternative river manage-ment schemes should be simulated first, and then carefully monitoredwhen the schemes are put into practice . In short, the capability ofsuccessfully managing a river for many uses depends on the capa-bility to predict effects .
The capability of predicting biological effects is particularly im-portant, because desirable functions of aquatic ecosystems, such aswaste assimilation and game fish production, depend on livingorganisms .
The standard fish bioassay, which uses death as a response,enables one to predict the toxicity of a particular waste to fish . Onelimitation of the standard bioassay is that it uses a grab samplewhich represents the quality of the waste at one point in time . Thewater used to make the dilutions is also taken at one point in time .At the actual industrial site, the quality of the waste and the riverwater vary through time. A composite waste sample partiallyovercomes this limitation, but may mask variations that are bio-logically important . For example, the toxicity of zinc to fish isaffected by the calcium concentration and temperature of the water
Received November 2, 1971 .
1 5 1
(CAIRNS & SCHEIER, 1957) . A fish kill might occur where the zincconcentration remained at a "safe" level, but there was a decreasein calcium concentration .
One could put fish in a continuous flow of waste diluted withriver water, but then there is one further limitation of the standardbioassay : death is used as the response . In order to prevent damageto organisms, it is necessary to have an early warning of dangerousconditions, so that corrective action can be taken . In other words,symptoms of ill health, which occur before death, must be detectedif there is to be time for diagnosis and treatment .
Techniques have been developed in our laboratory for detectingsymptomatic changes in the movement and breathing of fish . Thesymptoms occur early enough to permit survival of the fish if theintroduction of the toxicant (zinc in these experiments) is stoppedwhen the symptoms appear .
METHODS AND MATERIALS
Fish Movement PatternsFish movement patterns were monitored using the technique of
light beam interruption described in detail by CAIRNS et al. (1970) .Dawn and dusk were simulated in the experimental room by amotor-driven dimming unit which gradually increased the intensityof the room lights over a half-hour period starting at 6 : 30 a.m. andgradually decreased the intensity to 0 over a half-hour period star-ting at 6 : 30 p.m. The cumulative movement of each of six bluegillsunfish, a single fish per tank, was recorded every hour throughouta test except during the simulated sunrise and sunset when an addi-tional record was made on the half hour . Each day was divided intofour intervals ; first half day, second half day, first half night andsecond half night (Table I) . Before any statistical analysis could beperformed recordings for day 1 had to be completed . After thecumulative movement for day 1 was recorded statistical analyseswere performed after the completion of each designated time inter-val . For example, the cumulative movement recorded hourly foreach fish during day 1, first half day values was compared to thecumulative movement recorded hourly for each fish during day 2,first half day values . The statistical test used was a two sample testfor homogeneous variance (SOKAL & ROHLF, 1969) . If the statisticaltest indicated homogeneous variance a zero was scored (Table I)and the fish was considered to be exhibiting a normal movementpattern. If the statistical test indicated heterogeneous variance thefish was considered to be showing abnormal movement and an
152
asterisk was recorded (Table I) . As can be seen from the results ofthe experiment presented in Table I, the statistical test occasionallyindicated abnormal movement (day 1 vs. day 2, second half dayvalues, fish 6) during a period in which no toxicant was being addedto the system . However, a positive test for abnormal movement wasnever recorded for more than a single fish during a given timeinterval unless toxicant was being added to the system . Whenevera positive test for abnormal movement was scored the cumulativemovement recorded for the most recent time interval was droppedand the values recorded during the preceding interval were com-pared to the next recorded interval . For example, when the positivetest for abnormal movement was recorded for day 1 vs . day 2, fish 5,second half day values, the cumulative movement for day 2 wasdropped and where the table shows day 2 was compared to day 3for this fish during this interval, the actual comparison was betweenthe movement recorded for day 1 and day 3 .
Based on the results of 20 laboratory experiments "stress detec-tion" was defined as the presence of two or more abnormal move-ment patterns recorded during the same time interval .
A series of experiments at progressively lower measured zincconcentrations were used to determine the lowest concentration de-tectable by the movement apparatus .
Fish BreathingBreathing rates were determined from polygraph recordings of
breathing signals from 52 bluegill sunfish used in nine experiments .The fish were tested in plexiglas tubes through which dechlorinatedtap water or zinc solutions were metered at a flow rate of approxi-mately 100 ml/min . Breathing signals were detected by three plati-num wire electrodes placed in the water ; an active electrode, anindifferent electrode, and a ground . The test chambers and methodsof acclimating the fish are described in more detail by CAIRNS et al .(1970) . The photoperiod was the same as that for the fish movementstudy .
The fish were placed in test chambers by 6 : 00 p.m . and therecordings began at 6 : 00 a.m. the next day to allow the fish torecover overnight from handling. Zinc solutions were introduced at10 : 00 a.m. after the experimental fish had been exposed to tapwater containing no added zinc for periods of one to six days . Eachexperimental fish thus served as its own control . In addition, one ortwo fish were never exposed to zinc, and served as controls throug-hout each experiment. In one experiment, reported in Table VI,six control fish were exposed to water containing no added zinc forfour days .
1 53
Preliminary evidence suggested that the data could be analyzedby separating the experimental day into four periods ; a period from6 : 00 to 8 : 00 a.m. when the breathing rates changed markedly,a period from 9 : 00 a .m. to 5 : 00 p.m. when the rates were com-paratively high, another period of rapid change from 6 : 00 to 8 : 00p.m., and a night period from 9 : 00 p.m. to 5 : 00 a.m. when therates were comparatively low (SPARKS et al ., 1970) .
Bluegills increase their breathing rates when exposed to zinc(CAIRNS et al ., 1970) . Since a response is commonly considered to bea reaction to a stimulus, it seemed justifiable arbitrarily to define a"response" in these experiments as an increase in breathing rate .An individual fish was considered to have shown a response eachtime its breathing rate during a time period exceeded the maximumbreathing rate observed during the corresponding period of thefirst day, before any zinc was added. The maximum breathing rateof a particular fish during the dawn period of the first day (6 : 00 to8 : 00 a.m.) was compared to the breathing rates recorded for thatfish during the dawn period of the second day (Table IV) . A res-ponse was scored for each value on the second day that was higherthan the first day maximum . The same procedure was followed incomparing the maximum rate recorded during the 9 : 00 a.m. to5 : 00 p .m . period of the first day to the breathing rates recordedduring the same period of the second day . The dusk and nightperiods were compared in the same way . The same procedure wasfollowed in comparing the maxima of the first day to values recordedon the third day, the fourth day, etc . The rationale for this methodof analysis is as follows . The breathing rates during all periods of thefirst day were generally slightly higher than rates during comparableperiods of subsequent days, perhaps due to incomplete recoveryfrom the stress of handling . Any increase in breathing rate after thefirst day, that exceeded the maxima observed on the first day, couldthus reasonably be ascribed to some sort of stress . The control periods(before any zinc was added) and the experiment where no zinc wasadded at all were used to determine how many false detections thismethod of analysis would produce . The experimental periods (afterzinc was added) determined how quickly the method of analysiscould detect zinc concentrations in water .
Zinc concentrations were determined daily by atomic absorptionspectrophotometry .
RESULTS
Fish Movement PatternsTable I shows the results of one continuous flow experiment
154
First Half Day Values
Fish
1
0
0
0
0
0
0
0
0
0
0
0
0
0
*0
00
00
2-C
0
0
0
0
0
0
0
0
0
0
0
0
0
00
00
00
3
0
0
0
0
0
0
0
0
0
0
0
0
0
*0
00
00
4
0
0
0
0
0
0
0
0
0
0
0
0
0
00
00
00
5
0
0
0
*
0
+
*
0
0
0
0
*
0
*0
00
00
6
0
0
0
0
0
0
0
0
0
0
*
0
0
0*
*0
00
Seco
nd H
alf
Day
Valu
es
1
0
0
0
0
0b
0
0
0
0
0
0
*
0
**
*0
00
2-C
0
0
0
0
0b
0
0
0
0
0
*
0
0
00
00
00
3
0
0
0
0
0
0
0
0
0
0
0
0
00
00
00
4
0
0
0
0
0 + 0
*
0
0
0
0
0
0
00
00
00
5
0
0
0
0
0a
0
0
0
0
0
0
0
0
00
00
00
6
*
0
0
0
0N
0
0
0
0
0
0
0
0
00
00
00
Firs
t Ha
lf N
ight
Val
ues ---
------
------
---1
0
0
0
0
0
0
0
0
0
0
0
0
00
00
00
2-C
0
0
0
0
0
0
0
0
0
0
0.b
0
0
00
00
00
3
0
0
0
0
0
0
0
0
0
0
0 ¢ 0
00
00
00
4
0
0
0
0
0
0
0
0
0
0
0 + 0
0
00
00
00
5
0
0
0
0
0
0
0
0
0
0
0a
0
0
00
00
00
6
0
0
0
0
0
0
0
0
0
0
0N
0
0
00
00
00
Second Half Night Values
1
0
0
0
0
0
0
0
0
0
0
0
0
*
00
00
00
2-C
0
0
0
0
0
0
0
0
0
0
0
0
0
00
00
00
3
0
0
0
0
0
*
0
0
0
0
0
0
0
00
00
00
4
0
0
0
*
0
0
0
0
0
0
0
0
0
00
00
00
5
0
0
0
0
0
0
*
0
0
0
0
0
00
0*
00
6
0
0
0
0
0
0
0
0
0
0
0
0
0
**
*0
00
TABLE
ISt
atis
tica
l an
alys
is o
f li
ght
beam
int
erru
ptio
ns r
ecor
ded
duri
ng d
ays
1-20
of
cont
inuo
us f
low
expe
rime
nt 2
0.
Blue
gill
(Zn
++ a
ddit
ion
to s
tres
s de
tect
ion)
.
Day 1 Day 2 Day 3 Day 4 Day 5
Day 6 Da
y 7
Day
8 Da
y 9
Day
10 D
ay 1
1Day 12
Day 13
Day 14
Day
15 D
ay 1
6 Da
y 17
Day
18
Day
19vs
vs
vs
vs
vsvs
vs
vs
vs
vs
vsvs
vsvs
vs
vs
vs
vs
vsDay 2 Day 3 Day 4 Day 5 Day 6
Day 7 Day 8 Day 9 Day 10 Day 11 Day 12
Day 13
Day 14
Day 15
Day
16 D
ay 1
7 Da
y 18
Day
19
Day
20
carried out for 20 days. During this experiment fish were exposed to zinc on day 7 from 1 : 00 p.m. until 7 : 00 p.m. at which time the flow was returned to normal dilution water. The zinc concentra- tions reached their maximum at 7 : 00 p.m. and atomic absorption analyses on effluent samples collected at this time showed the fol- lowing concentrations : tank one, 13.32; tank two, less than 0.08; tank three, 11.39; tank four, 12.72; tank five, 13.32; and tank six, 12.59 mg/l Zn + +. The results show that these concentrations of zinc developing over the six hour interval of exposure were insufficient to cause a detectable change in the movement patterns of the fish. By 8 : 30 a.m. of day 8 the effluent zinc concentrations were less than 0.30 in all cases.
To determine the percent survival and recovery patterns of the fish once stress detection occurred, zinc flow was re-initiated at 1 : 00 p.m. on day 13 of this experiment. Between 8 : 00 and 9 : 00 p.m. on day 13 the zinc concentration in the effluent reached a maximum of: 7.51 for tank one; less than 0.05 for tank two; 7.49 for tank three; 7.52 for tank four; 7.49 for tank five; and 7.54 mg/l for tank six. The concentrations remained near the above values until the statistical analyses showed “stress detection” during the first half night values on day 14 (Table I). As soon as stress detection occurred the flow was returned to normal dilution water. At 10 : 00 a.m. on day 15 zinc analyses showed all effluent concentrations to be less than 0.70 mg/l Zn+ +. Stress detection continued to be registered for two consecutive time intervals following the initial detection, but after that no stress detection was registered and the frequency of abnormal patterns returned to pre-stress levels within 48 hours. In this experiment as with all others in which dilution water containing zinc was replaced with dilution water minus zinc at the time of stress detection all fish survived !
The results from the series of experiments at progressively lower zinc concentrations indicate that the lowest detectable concentra- tion is between 3.65 (Table II) and 2.93 mg/l zinc (Table III) for a 96-hour exposure.
Fish Breathing Table IV shows the breathing rates of five fish on days 1, 2, and
7 of experiment 8. The first four fish were exposed to a measured zinc concentration of 4.16 mg/l, beginning at 10 a.m. on day 7. The fifth fish served as a control and was not exposed to any added zinc. The amplitude of the breathing signals decreased every night, and the breathing rates for fish 2, in particular, could not be determined during some portions of the dark period (7 : 30 p.m. - 7 : 00 a.m.). The maximum breathing rates for each fish during each period of
156
TABLE II
Statistical analysis of light beam interruptions recorded during continuous flow experiment 16 . Bluegill4.2 mg/l Zn++ (3.65 mg/l Zn++) .
First Half Day Values
the first day are circled. The breathing rate of any fish during a timeperiod of day 2 or day 7, which is greater than the maximumbreathing rate recorded for that fish during the corresponding timeperiod of the first day has a rectangle drawn around it . The totalnumber of fish showing increased breathing is given at the bottomof each column. On day 2, fish 2 showed increased breathing onjust two occasions . In contrast after zinc was added on day 7, threeand four experimental fish at a time showed increased breathing .
Table V summarizes the results of successive comparisons of thefirst day maximal breathing rates to breathing rates on subsequent
157
Fish1234-C56
000000
000000
0000*0
00000(3vs5)
00*000
000 (5 vs 7)000
00*000
1 0 0Second Half Day Values
0 0 00 02 0 0 0 0 0 0 03 0 0 0 v 0 0 0 04-C 0 0 0 * * (4 vs 6) * (4 vs 7) * (4vs 8)5 0 0 0 c 0 0 0 06 0 0 o N 0 0 0 0
1 0 0First Half Night Values
0 0 00 02 0 0 * 0 (3 vs 5) * * (5 vs 7) * (5 vs 8)3 0 0 0 0 0 0 04-C 0 0 0 0 0 0 05 0 0 0 0 0 * 0 (6 vs 8)6 0 0 0 0 0 0 0
1 0 0Second Half Night Values
0 0 00 ! 02 0 0 0 * 0 (4vs6) 0 03 0 0 0 0 0 0 04-C 0 0 0 0 0 0 05 0 0 0 0 * 0(5vs7) 06 0 0 0 0 * * (5 vs 7) * (5 vs 8)
Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7Vs Vs Vs Vs Vs Vs Vs
Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8
TABLE IIIStatistical analysis of light beam interruptions recorded during continuous flow experiment 17. Bluegill
3.5 mg/l ,zn++ (2.93 mg/l Zn++) .
158
First Half Day Values
days (SCM method of analysis), for experiment 8 . During the con-trol period before any zinc was added there were 15 occasions whena single experimental fish responded, and three occasions when twoexperimental fish responded at the same time . At no time during thecontrol period did more than two fish show responses together .After the zinc was introduced, all four of the exposed fish showedresponses simultaneously on five occasions, and three fish showedresponses during the same time interval on 19 occasions . If thecriterion for detection of water conditions potentially harmful tofish were two or more responses during the same time period, then
Fish123-C456
000000
000000
00000*
0
00
0I
0
00
*0
00 (3 vs 5) 0
0000 (5 vs 7)00
000000
I 0 0Second Half Day Values
0 00
II
0
02 0 0 0
0
0 0 03-C 0 0 0 a 0
0 0 04 0 0 0
0
0 0 05 0 0 0 a
0
0 0 06 0 0 0 N 0
0 0 0
1 0 0First Half Night Values
0 00 0
02 0 0 0 0
0 0 03-C * 0 (1 vs 3) 0 0
0 0 04 0 0 0 0
0 0 05 0 0 0 0
0 0 06 0 0 0 0
0 0 0
1 0 0Second Half Night Values
0 00 1
0
02 0 0 0 *
0 (4 vs 6) 0 03-C 0 0 0 0
0 0 04 0 0 0 0
0 0 05 0 0 0 0
0 0 06 0 0 0 0
0 0 0
Day I Day 2 Day 3 Day 4 Day 5 Day 6 Day 7Vs Vs Vs Vs Vs Vs Vs
Day 2 Day 3 Day 4 Day 5 Day 6 Day 7 Day 8
TABLE IV
Breathing Rates of Bluegills, Experiment 8
Day
1
Note
: Bl
anks
ind
icat
e th
at t
he a
mpli
tude
of the breathing signal was so low that the rate could not be determi
ned
.+
Meas
ured zinc concentration of 4
.16
mg/l
int
rodu
ced,
exc
ept
for
Fish
5,
whic
hwas
not
exp
osed
to
zinc
.O Maximum breathing rate for each fish d
urin
g ea
ch p
erio
d of
the
fir
st d
ay.
p Br
eathing rates on second and seventh
days which exceeded first day maxima
. Th
e to
tal
numb
er o
f fi
sh s
howing increased breathing rates is shown
at t
he b
otto
m of
eac
h co
lumn
.
Peri
odHour
6Dawn
7:30
89
1011
12Light
12
34 5
66:30Du
sk 7 7
:30
8
910
1112
Dark 1
23
45
6:30
7
Fish
127
3039
®42
4042
3941
3738
(935
345©
2523
2022
2021
2524
2426
Fish
229
2029
344 32
2828
2827
4640 ©9
4940
2716
1215
14Fi
sh 3
1112
1519
1815
1818
1615
.16
1413
1610
911
1010
912
812
Fish
411
1116
1613
1614
1412
1112 12
1313
9
98
1010
78
810
Fish 5C
2121
2330®
3636
3532
3235
3233
342
37
2721
1920
1819
18160
1916
Day
2
Fish
120
2129
2732
3534
3026
3027
2233 22
2431
22
1515
1617
1415
1315
1613
16Fi
sh 2
1824
2928 40
4234
2630
2445 42
4622
19
2824
15©
1915
15Fi
sh 3
129
1012
1414
1512
1812
1213
1315
1616
13
1811
1010
1011
1213
1010
10Fi
sh 4
98
1410
911
1011
1010
1010
1013
1213
10
128
77
88
98
810
10Fish 5C
1718
1926
3330 37
3333
3130
3227 31
3628
29
2322
2117
1616
2016
1718
21
Tota
l0
00
00
0 0
00
00
00 0
00
0
00
00
01
01
00
0
Day
7
Fish
119
2012
1816
22 24
2615
28Ct
2226
3430
3830
5752
5048
4846
4743
4849
461
Fish
216
2831
3234 34
3624
2728
2820
4640
4042
1820
2228
23Fi
sh 3
1416
1716
14160
2016
1512
1315
1415
1615
1810
1612
914
1614
912
Fish
411
98
1011
1011
1216
1616
1111
1011
1012
1110
6261
59n5
4959
5a
Fish 5C
1616
1823
2428 28
2824
2530
2726
2629
2826
1115
1716
1414
1518
1515
15
Total
00
00
00
10
00
10
00
00
0
10
33
43
33
32
2
TABLE V
Number
offish
show
ing
resp
onse
s, b
efor
e an
d af
ter
expo
sure
to
4.1
6mg/l
zinc
.rn
4.Measured zinc concentration of 4
.16
mg/l
int
rodu
ced.
Resp
onse
s ob
tain
ed d
urin
g zi
nc e
xpos
ure
are
unde
rlin
ed.
Note
: There were 4 experimental fish (Ex) from Experiment 8 and I control fish (Con) from Experime
nt 8
.
Time
Day
6am
7 8
9 10
1112
lpm
23
45
67
89
10
1112
lam
2 3 4 5
2Ex
00 0
0 0
00
00
00
00
00
0 0
01
00 1 0 0
Con
00 0
0 0
00
00
00
00
00
0 0
00
00 0 0 0
3Ex
00 0
0 0
00
00
00
01
000 0 0
00
0 1 0 0
Con
00 0
0 0
00
00
00
00
00
0 0
00
00 0 0 0
4Ex
00 0
0 0
00
00
00
10
00
0 Re
cord
er off
-----
Con
00 0
0 0
00
00
00
00
00
-- _ _
1
-
5Ex
---
0 0
00
00
10
01
00
1
1
02
1
1
1
0Con
---
0 0
00
00
00
00
00
0 0
00
00 0 0 0
6Ex
Reco
rder
off
0 0
00
00
00
00
00
0
1
10
12 0 1 2
---
0 0
00
00
00
00
00
0 0
00
00 0 0 0
7Ex
00 0
0W
10
00
10
00
00
03 3
43
33 3 2 2
Con
00 0
0 0
00
00
00
00
00
0 0
00
00 0 0 0
8Ex
20 0
1
02
00
11
12
11
04 4 3
34
3 4 3 1
Con
00 0
0 0
00
00
00
00
00
0 0
00
00 0 0 0
9Ex
00
11
21
31
12
11
12
22
3
33
33 2 2 2
Con
00 0
0 0
00
00
00
00
00
0 0
00
00 0 0 0
10Ex
02 0
0 0
00
01
00
10
01
2
3
22
33
1 2 2
Con
00 0
0 0
00
00
00
00
00
0 0
00
00 0 0 0
11Ex
00 0
0 0
end
of e
xper
imen
t
Con
00 0
0 0
three false detections would have occurred before any zinc wasadded, and 4 .16 mg/l zinc would have been correctly detected eighthours after it was introduced . If the detection criterion were threeor more responses during the same time period, then no false detec-tions would have occurred and the zinc would still have been cor-rectly detected after eight hours .
The lowest zinc concentration tested was 2.55 mg/1. Using adetection criterion of simultaneous responses by three fish, thisconcentration was detected 52 hours after the zinc was added, withno false detections occurring during the four hours before zinc wasadded (Table VI) . The responses of six control fish that were ex-posed to tapwater containing no added zinc are also shown forcomparison. Note that there was no tendency toward increasedbreathing rates through time in the control fish, and that no morethan one control fish showed an increased breathing rate duringone time period .
Table VII summarizes information on three experiments thatindicates the effectiveness of the SCM method of analysis whendifferent criteria for detection are used . Changing the criterion fordetection from one to three responses per time period generallyincreases the lag time and decreases the number of false detections .The lag time is the time from the addition of zinc to the first detec-tion. A false detection is one occurring before any zinc is added tothe water .
DISCUSSION
The experiments described above show that the movements andbreathing rates of bluegill sunfish can be used to detect sublethalconcentrations of zinc. The criterion for detection is a certain num-ber of fish showing an arbitrarily defined response in breathing rateor activity during one time period .
In choosing a specific criterion for detection, the risk of notdetecting stressful conditions soon enough must be weighed againstthe risk of false detections, and the choice would probably be deter-mined by the nature of the pollutant . If a pollutant is easily detectedby the biological monitoring system, is slow-acting, and if the toxiceffects are reversible, then the criterion for detection might be re-sponses by 3/4 of the test fish, to avoid the false detections that wouldnecessitate expensive remedial action or a temporary shut-down .On the other hand, an industry that produces an effluent containinga fast-acting toxicant whose effects are irreversible would probablyuse a criterion that leads to rapid detection (responses by 1 /4 to 1 /2of the test fish), and would have to go to the expense of installing
161
TABLE VI
Number
offish
show
ing
resp
onse
s, b
efor
e an
d af
ter
expo
sure
to
2.5
5 mg
/l z
inc.
4.Me
asur
ed z
inc
conc
entr
atio
n of
2.5
5 mg
/l i
ntro
duce
d. Responses obtained duringzi
nc e
xpos
ure
are
unde
rlin
ed.
Note
: There were 3 experimental fish (Ex) from Experiment 7 and 6 control fish (Con) from Experime
nt 1
.
Time
Day
6am
78
910
1112
lpm
23
45
67
89
1011
12la
m2
34 5
2Ex
00
11
,I,1
11
11
11
12
12
11
11
01
11
1Con
00
00
00
00
00
00
00
00
10
00
00
0 0
3Ex
12
11
11
11
12
11
11
01
00
01
12
1
-Con
00
01
11
10
00
00
11
00
00
10
00
1
0
4Ex
Recorder off
11
00
23
32
10
01
12
12
22
22 3
Con
0
0 0
00
01
00
00
00
00
01
00
00
00 0
TABLE VIIEffectiveness of SCM method of analysis
(successive comparisons of maximal first-day breathing rates to rates on subsequent days) .Detection criterion :
holding ponds or recycling facilities to accomodate a relatively highnumber of false detections . Alternatively, a safety factor could beintroduced by metering proportionally more waste into the dilutionwater delivered to the test fish than is delivered to the stream . Thesafety factor could be determined by growth and reproductionexperiments with fish .
In an actual industrial situation water and waste qualities areapt to vary unpredictably, and it would certainly be desirable tohave a redundant detection system . It is conceivable that someharmful combination of environmental conditions and waste quali-ty would be detected by monitoring one biological function, butnot by monitoring another . It is also possible that excessive turbiditywould disrupt the light beams of the movement monitor, and notaffect the breathing monitor ; or that an excessive concentration ofelectrolytes would affect the electrodes of the breathing monitor andthe activity monitor have been combined in our laboratory forfurther experiments (Fig . 1) .
The rate of data acquisition and analysis could be greatly speededup if the monitoring system were automated as shown in Figure 2 .The sampling rate would be controlled by a minicomputer whichcould receive data from the movement monitor and the polygraphvia a multiplexer as often as every minute . The minicomputer wouldbe programmed to perform statistical analyses every 10 minutes,for example, and output the results on a teleprinter .
Figure 3 shows how the fish monitoring units would be used at anactual industrial site . A monitoring unit would be located on each
163
Experiment
Zinc
minimum no. ofConcentration No . of fish fish showing response
Lag time (Hoursfrom addition
of zinc)No. of falsedetections(mg/1) exposed at one time
9 5.22 3 1 0 12 in 100 hours2 4 tin 100 hours3 not detected after
45 hours0 in 100 hours
8 4.16 4 1 0 19 in 123 hours2 11 3 in 123 hours3 11 0 in 123 hours
7 2.55 3 1 0 2 in 4 hours2 8 0 in 4 hours3 52 0 in 4 hours
TO POLYGRAPH
ELECTRODE
TO COUNTERFig . 1 . Test chamber for monitiring system, showing the electrodes for recordingfish breathing and the light-beam system for recording fish movement .
1 64
18 CHANNELS
MULTIPLEXER
MINI COMPUTER
MOVEMENTMONITOR
6 CHANNELS
POLYGRAPH
Fig. 2 . An automated fish monitoring unit .
waste stream in the plant and on the combined waste stream . Theexperimental fish in each unit would be exposed to waste dilutedwith water from the river above the plant, and control fish would beexposed to upstream water alone (Fig . 4) . The information fromeach monitoring unit could be analyzed by a central data processor,
TELEPRINTER
TANKS
MINNIMMMWASTE WASTE
2
00
Fig . 3. Arrangement of fish monitoring units at an industrial site .
and when there was a warning response, the industry could tellwhich waste stream was at fault . If the problem was outside theplant, the control fish would show responses .
Figure 5 shows how the in-plant monitoring systems would beintegrated into a river management system. The in-plant monitoringunits are shown as squares, and in addition to supplying informa-tion to each industry, the monitoring units also inform the controlcenter. In such a system, there are several alternative damageprevention measures that could be used, in addition to whatevermeasures, such as shunting wastes to a holding pond or recyclingwastes for further treatment, are available to each industry . If themonitoring units at Industry 2 indicate that toxic waste conditionsare developing, then the control center might have Industry 1 holdits waste until the danger of combining wastes from Industry 1and 2 in the river were alleviated by control measures at Industry 2 .Alternatively, the control center might call for a release of waterfrom the upstream dam to dilute the effluent from Industry .
It is likely that "fish sensors" in continuous monitoring units atindustrial sites can warn of developing toxic conditions in time toforestall acute damage to the fish populations in streams . In con-junction with stream water quality standards for chronic exposure,such biological monitoring systems should make it possible forhealthy fish populations to coexist with industrial water use .
165
UPSTREAMWATER
eMOVEMENT
MONITOR
BREATHING
MONITOR
IN - PLANT MONITORING UNIT
WASTE
TO CENTRALPROCESSOR
DRAIN
Fig. 4. Detail of a single fish monitoring unit, showing how the experimental fishare exposed to waste diluted with upstream water and the control fish are exposedto upstream water alone .
ACKNOWLEDGEMENT
This research was supported by grants 18050 EDP and 18050EDQ from the Water Quality Office, Environmental ProtectionAgency, and by funds provided by the United States Departmentof the Interior, Office of Water Resources Research, administeredby the Water Resources Research Center as project A-039-VA . Wethank the personnel of the Mc Kinney Lake National Fish Hatchery,Hoffman, North Carolina, for supplying the fish used in the ex-periments on fish breathing .
166
O
4
O
4
O 0 0 O 4
P 4 4 OO
4
O
4
O O OO
OWOODED AREA
CONTROLO p
4 O
4
4
CENTER
O 4
a RESERVOIR Y
4
MIXED INDUSTRIAL AND AGRICULTURAL AREA
/.-
I
nMONSON
Q
SEWAGEP ANT
.~
RESERVOIR X
O
V
NDUSTRY1
OUSTR2
U
4 4 4
00000000000000000000000000000000000000000000000000
NDUSTRY3
∎
90009990SAGA9 0 9 90 G 0 99 9 9 999999 9 9
4
O
O
4
AGRICULTURALAREA
000000000000000000000000
9909999900900 9 0 9 0 90099001
4
O
O `f O O
40 4
O O
O O
P
4N-PLANT MONITORING UNIT
P
4
4
4
P
d IN-STREAM MONITORING UNITA
Fig. 5 . Use of in-plant monitoring units in a river management system .
REFERENCES
BRUNGS, W. A. - 1969 - Chronic toxicity of zinc to the fathead minnow, Pimephalespromelas RAFINESQUE . Trans. Amer. Fish. Soc . 98 (2) : 272-279 .
CAIRNS, J ., JR., DICKSON, K. L., SPARKS, R. E. & WALLER, W. T . - 1970 - Apreliminary report on rapid biological information systems for waterpollution control . ,q. Water Poll. Contr. Fed. 42 (5) : 685-703 .
EATON, S . G. - 1970 - Chronic malathion toxicity to the bluegill (Lepomis macro-chirus RAFINESQUE) . Water Research . 4 : 673-684.
MCKiM, J. M. & BENOIT, D . A. - 1971 - Effects of long-term exposures to cooperon survival, growth, and reproduction of brook trout (Salvelinus fontinalis)Y. Fish. Res . Bd. Canada 28: 655-662 .
MOORE, J. G., JR., Commissioner . - 1968 - Water Quality Criteria . Report of theNational Technical Advisory Committee to the Secretary of the Interior .U.S. Govt. Printing Office. 234 pp .
MOUNT, D . I . - 1968 - Chronic toxicity of copper to fathead minnows (Pimephalespromelas RAFINESQUE) . Water Research . 2 : 215-223 .
MOUNT, D. I . & STEPHAN, C . E . - 1967 - A method for establishing acceptabletoxicant limits for fish-Malathion and the butoxyethanol ester of 2, 4-D .Trans. Amer . Fish . Soc. 96 (2) : 185-193 .
SOKAL, R. R. & R0HLF, F . J . - 1969 - Biometry. W. H . Freeman and Co . 776 pp .SPARKS, R. E., WALLER, W . T ., CAIRNS, J . JR . & HEATH, A . G. - 1970 - Diurnal
variation in the behavior and physiology of bluegills (Lepomis macrochirusRAFINESQUE) . The ASB Bull . 17 (3) : 90. (Abstract) .
SPRAQUE, J . B . - 1969 - Measurement of pollutant toxicity to fish I . Bioassaymethods for acute toxicity . Water Research . 3 : 793-821 .
167