United States
Environmental Protection
Agency
Environmental Research
Laboratory
Duluth MN 55804
Research and Development
EPA/600/S3-89/011 Aug. 1989
Project Summary
An Automated Monitoring
System for Fish Physiology and
Toxicology
Richard W. Carlson, Gregory J. Lien, and Bruce A. Holman
This full report describes a data
acquisition and control (DAC) system
that was constructed to manage
selected physiological measure-
ments and sample control for aquatic
physiology and toxicology. Auto-
mated DAC was accomplished with a
microcomputer running menu-driven
software developed with an extended
BASIC. An Interface module was built
that connected standard sensors and
controls to the computer. Digital I/O
signals for sample device control and
analog signals from sensors were
multiplexed through the interface
module. Time intervals for automated
DAC were user defined, and test data
were displayed on a monitor, printed,
stored on disk, and transferred to a
minicomputer for analysis. Auto-
mated measurements were made of
temperature, ventilation volume, oxy-
gen content of exposure (Inspired)
and expired water, and pH of both
waters from four in vivo rainbow trout
Salmo gairdneri preparations. Oxygen
uptake efficiency and oxygen con-
sumption were calculated. Urine and
expired water samples were also
collected from all fish.
Non-automated sampling included
ventilation frequency, cough frequen-
cy, the electrocardiogram, and aortic
blood from an implanted canula.
Sampled blood was analyzed for
oxygen, carbon dioxide, pH, hemato-
crit, and hemoglobin. The respiratory-
cardiovascular data gathered with
this system were used to define fish
acute toxicity syndromes (FATS)
specific to known modes of toxic
action.
This Project Summary was devel-
oped by EPA's Environmental Re-
search Laboratory, Duluth, MN. to
announce key findings of the research
project that Is fully documented in a
separate report of the same title (see
Project Report ordering information at
back).
Introduction
One approach to understanding causal
relationships between chemicals and
their effects was developed recently and
is termed fish acute toxicity syndromes
(FATS). These are collections of direct
and indirect measures of effect, or clinical
signs, manifested in the animal upon
exposure to chemicals that are uniqi-e
and specific to a common mode of
action. Based on a group of measurable
toxic signs involving the respiratory-
cardiovascular system in rainbow trout
Salmo gairdneri, FATS have been de-
fined for narcotics, oxidative phosphor-
ylation uncouplers, acetylcholinesterase
(AChE) inhibitors, respiratory membrane
irritants, and the pyrethroid insecticide
fenvalerate.
FATS testing required data acquisition
on 11 respiratory-cardiovascular variables
and the capability to monitor more if
necessary. Monitoring was performed
manually during all previous FATS tests
and consumed the full attention of at
least three people along with the part-
time help of several others during both a
seven hour control period and for up to
48-h during the acutely lethal exposure
period. Only two fish could be prepared
and tested at one time, and two tests
were therefore required to gather suf-
-------
ficient information on four fish to reliably
define a FATS. Although measurements
were made often enough so that
statistical relevance was established, a
higher sampling frequency was desirable
for greater confidence.
The objective here was to develop an
automated system to efficiently quantify
some of the physiological functions of a
whole fish preparation exposed to acutely
lethal chemical concentrations. It was
desired that the system provide for data
gathering on at least four fish, and to do
so at predetermined intervals throughout
the test including periods of unattended
operation. Another requirement was that
it remain flexible enough so that sampling
and measurement regimes could be
changed, singly or collectively, during a
test. With these in mind, a system was
designed that could be constructed from
commercially available sensors, control
valves, a personal computer, a specially
constructed interface, and menu-driven
software to coordinate system activities.
Procedures
The system performed physiological
monitoring on rainbow trout Salmo
gairdneri that weighed between 0.6 and
1.0kg and were exposed to a lethal
concentration of an organic chemical.
This required the integration of several
subsystems that included: (1) exposure
apparatus that provided water and
toxicant delivery; (2) automated sampling
and measurement circuits and devices
that provided automatic data collection
and sampling of physiological functions;
(3) non-automated circuits and devices
that provided monitoring of those
physiological functions that defied auto-
mation at this time; (4) a microcomputer
system that controlled all aspects of
automated monitoring; and (5) an inter-
face that provided all necessary intercon-
nections and switching between the
computer and external devices.
The exposure system was the same as
that described by McKim and Goeden
(1982) except that the toxicant delivery
apparatus and fish chambers were
enclosed in a specially constructed and
vented enclosure to minimize human
exposure to potentially hazardous chem-
icals. Each of the four respirometer-
metabolism chambers (Figure 1) was
modified so that water overflow from the
A compartment on fish chamber one and
from all B compartment standpipes was
directed into flow measuring devices;
other A compartment and all C compart-
ment overflow went directly to drain.
From ports located on the sides of the
chambers, water was directed to the
different sensing electrodes without
aeration.
The interface fulfilled three important
needs. First, information in the form of
analog signals, or varying voltages, from
the meters measuring dissolved oxygen
concentration, pH, temperature, and from
the pressure transducers for water flow
rate must be read into the computer.
Secondly, the computer was required
to control the operation of solenoids and
motors used in making flow measure-
ments and water samples. Additionally,
the toxicant pumps were controlled and
the urine fraction collector was advanced
through the interface interconnections.
Lastly, the interface provided power to
operate or control the operation of the
different sensors and devices attached to
it. An internal power supply provided +
12, + 15, and -15 volts DC to the
system, while connections to power
supplies external to the interface supplied
+ 5, -5, and + 24 volts DC.
The microcomputer was an IBM PC/XT
specially manufactured for Analog
Devices and included Concurrent CP/M-
86 as its operating system, a multi-
tasking system that managed the I/O of
all devices attached to the computer,
provided file management, and loaded
and ran the operational program for DAC.
The operational DAC program was written
in-house and named "TEST." TEST
contained 1083 lines of source code and
required 55 KB of memory for the un-
documented source code or 48 KB for
compiled object code. TEST consisted of
a short main program to begin and direct
program execution, a timing task and an
interrupt task running concurrently, and
25 subroutines that performed all the
functions required by the main and task
portions of the program.
Figure 2 shows the overall layout of the
data acquisition and control (DAC)
system that performed sampling, meas-
urement, and calculation of selected
physiological functions. Individual compo-
nents and their operation are discussed
in the research report. The DAC system
monitored pH, dissolved oxygen (DO),
temperature, and flow rate of both the
incoming and expired water in which the
fish resided during a test. This was
accomplished by monitoring the expired
water (B compartment) of up to four fish
chambers and the incoming water (A
compartment) of fish chamber one
.(Figure 1) Additionally, samples of both
waters and urine fractions from each fish
were collected automatically and held for
chemical analysis. A single water sample
was taken from the A compartment when-
ever any or all of the B compartmer
were so scheduled. Also, whenever a
fish chamber was monitored for pH, C
or temperature, the A compartment w
sampled immediately afterward so th
the samples of inspired and expin
water were as close in time as possib
This was necessary because U
calculations for oxygen uptake efficient
(UE) and oxygen consumption (VO
involved the difference in DO content
both waters at that moment.
After each measurement or samplir
operation the computer monitor scree
was updated to show the results for th
time interval, a continuous hard copy w;
appended and all data were appended
a file residing on the hard disk.
During each FATS experiment mea
urements were made on the physiologic
variables shown in Table 1. Ventilatk
volume (VG), V02, and UE wei
monitored automatically while the remaii
der were done manually. Ventilatoi
frequency (fv) and cough frequency (1
were determined from portions of stri|
chart recordings made of the tro
ventilatory patterns. These were moi
itored from non-contact stainless ste
wire electrodes placed in the B and
compartments of each fish chamber.
Spinally-transected rainbow trout wei
each fitted with a latex rubber membrar
that separated expired water from incon
ing water, a dorsal aortic cannula f<
blood sampling, copper wire electrode
for monitoring the EKG, and a urinai
catheter. After surgery the fish wet
placed in individual respirometer chan
bers, the electrode connections madi
and the urinary catheter was connecte
to the C compartment port.
Results and Discussion
To date 17 tests involving 68 fish hav
been completed using the systerr
Testing included three freshwater contn
runs, two control tests on a carrier solver
used to aid dissolution of some te:
chemicals, and 13 tests with organi
chemicals used in describing fish acut
toxicity syndromes (FATS). Collectively
the results from these tests showed th;
the system performed as designe
despite some sporadic electronic ma
functions and problems with sensor cal
bration during some tests. The fir:
chemical tested with the automate
system, 2,4-dinitrophenol, was in th
group of uncouplers originally tested, an
the results obtained using the automate
system were consistent with thos
obtained manually. This verified that th
automated system was suitable for FAT
-------
II
Exposure
Water
Fraction
Collector
Toxicant
Pump
Notes:
Flow
Meter
li
Water
Sample
Flow
Meter
II
Water
Sample
1. Chamber design 81 surgical preparation (McKim &
Goeden. 1982)
2. Transected 6OO-1100 gram Rainbow trout fSalmo
gaidneriy
3. A Chamber (exposure water) monitored on only 1of4 fish
B Chamber (expired water) (Vg) ventilation volume
C Chamber (fresh water) input for maintaining temp
Figure 1. Schematic diagram of respirometer-metabolism chamber. Connections for
ventilatory pattern andEKG are not known.
testing as well as showing that the
responses used to define a FATS were
reproducible.
The main advantage gained by using
the automated system for FATS testing
was that four fish could be monitored
simultaneously for their cardio-respiratory
responses. When done by the automated
system, measurements of VQ and DO
were rendered effortless and it became
possible to monitor more than two fish
per test. This at least doubled the
number of FATS tests that could be done
in the same time frame.
Also, round-the-clock monitoring was
now possible and this greatly increased
the number of measurements done for
VG, UE, and O2. This ensured data
gathering throughout a test for every fish
and increased the confidence that data
were not missed for periods of critical
change.
Another advantage to using the
automated system was that certain
judgments concerning the course of an
experiment could be made while it was in
prcgress. For instance, it is characteristic
of narcosis-inducing chemicals that their
effects on an organism are reversible
even at the point of apparent death,
usually defined as respiratory arrest in
aquatic toxicology, whereas effects
induced by chemicals with more specific
modes of toxic action are irreversible. By
following VG and VO2 on the computer
printout as well as locomotor activity,
ventilation, and the EKG on those
recordings, the fish could be revived at
various stages of intoxication with
toxicant-free water and recovery
monitored if it were necessary to demon-
strate that certain chemicals were
narcotic.
Conclusions
1. Automated monitoring of respiratory-
cardiovascular variables from fish
resulted in a considerable savings of
time and effort when compared to
manual data gathering methods.
2. Automated monitoring provided
continual data collection during per-
iods of unattended operation, thus
ensuring that data were collected
during times when critical changes
may have occurred.
3. The real-time sampling and
calculation of vital signs permitted
judgments on the course of an
experiment.
4. Automated monitoring allowed rapid
data collection at shorter time
intervals than manually possible. A
greater number of samples provided
for greater statistical reliability.
5. Data were easily manipulated and
transferred between computers be-
cause they were immediately stored
in computer files.
6. Less manual sampling reduced
human exposure to potentially
hazardous chemicals.
References
McKim, J.M. and H.M. Goeden. 1982. A
Direct Measure of the Uptake
Efficiency of a Xenobiotic Chemical
Across the Gills of Brook Trout
(Salvelinus fontinalis) Under Normoxic
and Hypoxic Conditions. Comp. Bio-
chem. Physiol. 72C: 65.
-------
Table 1
Physiological Variables Monitored in Rainbow Trout to
Define the Toxic Responses Associated with Fish Acute
Toxicity Syndromes (FATS)
Variable
Units
Ventilation Volume (VQ)
Total Oxygen Consumption (V(>2)
(UE)
(fy)
Gill Oxygen Uptake Efficiency
Ventilation Frequency
Cough Frequency (fc)
Heart Frequency (fa)
Total Blood Oxygen (arterial) (TaO2)
Total Blood Carbon Dioxide (arterial) (TaCOz
Blood pH (arterial) (pHa)
Hematocnt (Hct)
Hemoglobin (Hb)
ml/min
mg/kg/h
%
no./min
no./min
no.lmm
glHX) mi.
mmol.L
pH units
%
gliOOmL
-------
System Configuration
DO
Analog
Mux
Digital
Mux
1
LJ
LJ
1
_J LLJ L£J L±J L*J
Temperature A
II 2 || 3 || 4 || 5 |
Temperature B
Flow Pressure Sensors
A/D
Card
Macsym
120
IBM XT
Computer
DIO
Card
f low Solenoids
'
~i H
1 1
i
7 1 1 •
> 1 1 .
1 1 1 C
Urine
Collector
Toxicant
Pumps
Water Sampler Multiplexer
IVarer Sample Valves
Figure 2. Block diagram of the automated system.
-------
The EPA authors, Richard W. Carlson (also the EPA Project Officer, see below),
Gregory J. Lien, and Bruce A. Ho/man, are with the Environmental Research
Laboratory, Duluth, MN 55804.
The complete report, entitled "
An Automated Monitoring System for Fish Physiology and Toxicology," (Order No.
PB 89-155 2121 AS; Cost: $15.95, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Research Laboratory
U.S. Environmental Protection Agency
Duluth, MN 55804
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S3-89/011
------- |