Introduction
Developmental aspects of the Monti-
cello Remote Sensing System (MRSS)
discussed in the full report are: the
system hardware configuration together
with the software program implementa-
tion, the characteristics of MRSS in
operation, and an evaluation of the data
from the field testing of the system which
took place during May 1979. Hardware,
software, and the microprocessor are
discussed in the engineering design
section. The results and discussion sec-
tions focus on the operational character-
istics of MRSS, examining the accuracy
of position information obtained with the
system, and also the experimental dif-
ficulties with respect to external condi-
tions at Monticello.
A 28-day performance test using 11
walleyes, Stizostedion vitreum vkreum,
with surgically implanted radio-transmit-
ters provided baseline data on the opera-
tion of MRSS and guidelines for interpret-
ing the position information recorded.
Manual testing of accuracy was also done
to verify that both the hardware and
software were operating according to
specification. To continuously monitor
reliability, transmitters in fixed, known
locations were used as references. Also
monitored were four "open channels,"
i.e., frequencies without actual transmit-
ters, but within the range of the operating
frequencies. These were used to deter-

-------
mine whether interference from external
electrical noise could disrupt the opera-
tion of MRSS.
Study Area
  The Monticello field station of the EPA,
Environmental Research Laboratory (Du-
luth, Minnesota) consists of eight exper-
imental channels, each 487 meters  in
length and composed of alternating pools
and riffles. The average dimensions of a
pool are 33  meters long by 4.0 meters
wide by 0.6  meters deep; the riffles are
approximately 33  meters long by 2.6
meters wide by .33 meters deep. Miss-
issippi River water is pumped in at one
end of each channel and returned to the
river through a culvert at the lower end of
the channel. Flow rates, pH, and water
temperature can  be controlled experi-
mentally; light intensity can be manip-
ulated by installing shades or covers over
part or all of  the channel system.
Conclusions
  The system (MRSS) designed and built
under this contract determined the loca-
tion of radio-tagged fish to the nearest
pool or riffle area by using a microproces-
sor to control all functions of the system,
keeping track of the time-of-day, printing
location data,  and  determining  signal
quality. The system was low cost, easy to
install, and used readily available com-
ponents. Sample rates and other param-
eters were easily  changed via key-
board input. Using  this system, a  re-
searcher could collect data without  dis-
turbing the  fish.  It was  designed  for
unattended operation to allow continuous
collection of data.
  The results  of the  performance  test
conducted with this  automatic radio
tracking system indicated that fish could
be located to the nearest pool or riffle area
at the Monticello Ecological  Research
Station (MERS) with  an  accuracy of
98.7%. Open  channels were used to
monitor background interference,  and
reference transmitters continuously ver-
ified system reliability throughout  the
experiment.
  The field test indicated that the data
produced  by the  tracking system was
sufficiently comprehensive to determine
changes in walleye behavior with respect
to external influences. The system  pro-
vided enough data on individual fish to
detect changes in movement and resting
patterns  in  addition to demonstrating
overall location preferences.
Recommendations
  Fish behavior relative to environmental
alterations can be studied with the auto-
matic radio tracking system developed for
this contract. Experimental stream chan-
nels at the MERS approximate a running
water environment; thus, data obtained
should be appropriate to questions con-
cerning stream  or river ecology  and
information collected from the channels
should substantiate laboratory data. The
radio  tracking system developed  and
tested under this  contract  will permit
researchers to observe mortality, avoid-
ance  behavior, and changes in  activity
patterns resulting from various toxicant
discharge schemes  and concentrations.
Fish behavior can be a timely and sens-
itive indicator of environmental disturb-
ance.  Experimentation conducted at the
MERS showed this system  can  provide
quantitative data for evaluation of water
quality criteria.
Engineering Design

Hardware and Electronics
  The goals were to design  a system
which would be able to locate individual
fish swimming in experimental channels
and  to  determine  their responses to
factors such as temperature, chemical
and pH  variations. The design require-
ments specified a minimum capacity of
30 fish per tracking cycle.
  Since design requirements were to
locate fish to the nearest pool or riffle in
the upper 360 m  of a given channel, rf
tags were used, and the signal strength
was measured at an antenna placed at
each pool and riffle area. Rf tags were
chosen because the pools often had dense
stocks of aquatic  vegetation, which im-
peded the transmission of sonic signals,
while the  fresh  water allowed good
transmission of an rf signal. The antenna
chosen was a horizontal dipole with the
axis of the antenna parallel to the channel
(Figure 1). The antennas were positioned
adjacent to the channel and  mounted
approximately one meter above the sur-
face. The  dipole design  was  chosen
because of  its  optimum field strength
versus distance characteristic, inexpen-
sive implementation, and long-term sta-
bility. Several other antenna designs were
tested  including  vertical dipoles, loop
antennas, and horizontally polarized di-
poles placed in the  water. Although an
immersed  horizontal  antenna actually
had better field strength versus distance
characteristics, problems with vegetation
removal, and other activities in the chan-
nel made using an above-water antenna
preferable.
  The  antennas were connected  to  a
relay control box  which switched one
antenna at a time into the main signal
cable (Figure 1). The cable chosen for the
main signal cable was a low-loss coaxial
type used for closed circuit television
systems. The  cable was  designed for
direct burial, and  its  low cost made  it
acceptable for use  in this application. To
help equalize signal attenuation, the
signal was taken off  the center  of the
main cable rather than the end (Figure 1).
Thus, the cable was effectively divided in
half, reducing signal differences to ap-
proximately  3db rather than  6db if  a
single length had  been used. A signal
amplifier was also used at this point to
maintain the signal-to-noise ratio and
compensate for losses as the signal was
fed back to  the measurement site. An-
tenna switching was controlled by a four-
bit binary select code with each of the
switch boxes containing a 4  to 16 line
decoder, relay  driver, and  a relay  to
connect each antenna to the signal cable
as directed by the four bit common from
the microprocessor.
  The receiver  used in this  application
was a standard memory receiver designed
by the Cedar Creek Electronics Labora-
tory. This receiver had 64 channels  of
digital  memory which allowed prepro-
gramming of transmitter frequencies.
These pre-selected frequencies could be
recalled by selecting 1 of the 64 memory
locations. The signal from the receiver
was detected and fed through a low pass
filter to  an  eight-bit  analog to  digital
conversion limit for measurement. Data
were recorded on a Date! thermal printer,
model number  DPP-Q7 (Figure 2). Day
number and time  of day were derived
from a hardware binary minute counter
which was used to count elapsed time in
order to update the system clock with the
entire  system  operation.  Timing  was
under microprocessor control.

Microprocessor Hardware
  Functions  of  the microprocessor in-
clude timing, control of events such as
issuing  frequency (fish)  select  codes,
switching antennas, and  initiating A/D
conversions. A CMOS RCA 1802  micro-
processor* was chosen for this applica-
tion because CMOS circuitry was compat-
ible with the circuitry in the Cedar Creek
'Mention of trademarks or commercial products does
 not constitute endorsement or recommendation for
 use

-------
                     Experimental Channel Schematic
                                                         Main
                                                        Signal
                                                         Cable
                                                            Receiver &
                                                          Microprocessor
Figure  1.   Diagrammatic representation of an experimental channel at the E.P.A. field station,
           Monticello, Minnesota.
Receiver
Channel-

Day
Number
Figure 2.
                 Antenna
                   with
                 Maximum
                 Signal

                 ^ Index of
                  Signal
                  (Signal—
                  Noise)
                 Time of day
                  'Hours,
                  Minutes
Data recorded by the Monticello
remote sensing system on ther-
mal paper.
receiver and had inherent noise immun-
ity. Its low power consumption was also
desirable for applications requiring bat-
tery  power.  System  configuration is
shown in block diagram form in Figure 3.
System control and programming was
done via a microterminal notepad sup-
plied by RCA for the 1802 system. The
standard utility ROM UT5 developed by
RCA was used  to support the micro-
terminal operation.
  Three pages of memory were provided
using CDP 1822 static RAM chips. These
chips had a 256 x 4  configuration  re-
quiring six chips to generate three pages.
The memory was protected by a standby
power source and was laid out for easy
substitution of  ROM chips for program
storage if desired. A National Semicon-
ductor ADC0808 eight-bit  binary A/D
converter was used for signal measure-
ment.

Software Description
  Software formed the most significant
portion of the research and development
for this project. Software methods were
emphasized rather  than the hardware
materials currently employed in  most
automated animal monitoring systems
because  of the flexibility afforded by
programming. This flexibility  proved val-
uable in overcoming technical difficulties
during the  development phase of the
project and should provide optimum
operating capabilities for the changing
requirements of future studies. Twelve
distinct software entities, each perform-
ing  a  particular task,  comprised the
program  for this system. An important
function  developed specifically  for the
Monticello area was a measure of signal-
to-noise ratio to determine the reliability
of each fish location.
  To facilitate software development, the
COSMAC Development System was pur-
chased from RCA. This development
system allowed programs to be written in
RCA COSMAC  assembly language, and
was designed to be used on a time sharing
system as a cross-assembler or debugger.
Assembly language allowed programs to
be written and modified using convenient
symbols rather  than machine language.
Using the RCA support package and the
University of Minnesota Cyber 74 NOS
time sharing facility, assembly language
source code was  converted into the
hexadecimal machine codes.  These ma-
chine instructions were then entered into
the RCA  1802  microprocessor memory
via the microterminal notepad.
  Initialization of all operating registers
occurred at the beginning of each cycle of
the main program. Each pass consisted of
a scan of all  12 antennas for a particular
fish  frequency  out of the 32 possible
channels. The  antennas were scanned
beginning with  number  13, which was
most distant from the receiver site, and
ending with the antenna located at station
2.
  At the end of each pass, the number of
the antenna with the maximum reading
was stored in the printer output buffer.
Next, the microprocessor compared the
maximum signal reading from the anten-
na nearest a fish to an average back-
ground noise reading from the remaining
12  antennas.  This important feature

-------
MRSS Hardware Configuration
Cedar
Creek
Programmable
Receiver
Page 1
Page 2
8200-82 IF
in the experimental channel. Movement
patterns of the walleyes indicated a
pronounced crepuscular activity pattern.
Figure 4 summarizes the location informa-
tion recorded by MRSS for one radio-
tagged walleye during the system evalua-
tion. On the x axis, each unit represents a
24-hour time period defined from noon to
noon, the y axis represents day of the
month. The contour lines depict the
percentage of location determinations
that occurred at each antenna during
each period, i.e., usage of each individual
pool or riffle by each walleye over time.
Each walleye favored one pool on a given
day or series of consecutive days. Larger
fish seemed to move away from the
release pool sooner than smaller fish.
Examination of the data for the two largest
fish revealed their utilization of mutually
exclusive primary pools for each time
period. Examination of movement pat-
terns indicated that the walleyes tended
to form heterogeneous-sized groups.
In summary, the Monticello Automatic
Fish Tracking system performed up to
specificat' , and can be used to observe
fish movement patterns, mortality, activ-
ity patterns, and social interactions. Alter-
ations in these behavioral parameters
should provide timely and sensitive
measurements of the effects of aquatic
toxicants.
Figure 3. The Monticello remote sensing system hardware device organization and informa-
tion flow.
permitted an index of signal reliability so
that marginal signals from fish or spur-
ious r.f. signals from nearby power lines,
unshielded ignition systems, and thunder-
storms could be eliminated. After results
of a pass were printed, the microproces-
sor either branched back to look at
another fish or, if the fish frequency had
been processed, waited a predetermined
time period (an installation parameter)
before beginning another series of obser-
vations for the fish frequencies in the
tracking queue.

Field Testing

System Verification
A 28-day system check was conducted
during May 1979. Eleven walleyes were
released and monitored on the automatic
tracking system. Four transmitters placed
at known locations in pools and riffles
were used as reference transmitters.
Periodically, walleyes were simultaneous-
ly monitored with manual locating equip-
ment to verify the tracking system. The
system made 5297 location determina-
tions on transmitters at known locations.
A total of 5229 (98.7%) of these locations
were positioned correctly by the tracking
system. Of the 68 errors that occurred, 42
(62%) occurred at poor signal-to-noise
levels and likely would have been rejected
during actual fish tracking operations.

Walleye Tracking
Over a 28-day period, 21,039 location
observations were made on 10 walleyes

-------
Antenna
Stations
Day Calculated from Noon to Noon

10 12 14 16 18 20
10-
12-
Figure 4. Contour map of the percent of location determinations in a 24-hour time period occurring in each pool or riffle for walleye 4041.
Larry Kuechle and Richard Reich/e are with the University of Minnesota, Cedar
Creek Bioelectronic Laboratory, Bethel, MN 55005; Donald Siniff is with the
University of Minnesota, Department of Ecology and Behavioral Biology,
Minneapolis, MN 55455; Kathlean Zinnel and Jon Ross are with the Cedar
Creek Laboratory and the Department of Ecology and Behavioral Biology at the
University of Minnesota; and Charles Kleiner (also the EPA Project Officer, see
below), is with the Environmental Research Laboratory, Duluth, MN 55804.
The complete report, entitled "Automatic Radio Tracking of Fish in Experimental
Channels," (Order No. PB 86-131216/A S; Cost: $ 16.95, 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
6201 Congdon Blvd.
Duluth, MN 55804
U. S. GOVERNMENT PRINTING OFFICE: 1986/646 116/20817

-------
United States
Environmental Protection
Agency
Official Business
Penalty for Private Use $300

EPA/600/S3-85/075
Center for Environmental Research
Information
Cincinnati OH 45268
*~> ^ ^ t i


\ -D.'",' v
1  ';:
J  - ,'-,, "r^
               0000329    PS

               U  S  ENVIR PROTECTION  AGENCY
               REGION  5  LIBRARY
               230  S  DEARBORN  STREET
               CHICAGO                IL   60604

-------