Íslenskar landbúnaðarrannsóknir - 01.09.1978, Qupperneq 63

DESIGN OF A SALMON COUNTER 61
Some operational data.
Sensor units are normally 10 m wide and
about 1.5 m long (the dimensions refer to
the river). Several units can be connected
to cover a river of over 10 m in with. This
simple type of sensor is most suitable for
rivers without a fish ladder or weir, where
tunnel sensors cannot be used. Optimal
water velocity is consider^d to be 0.5-1.5
m/s and optimal water depth lies between
0.4 and 0.8 m. Energized from a conven-
tional 12-volt storage battery ofabout 100
amperehours, the unit will operate up to
one month before it needs recharging. Re-
solution, e. i., the ability to distinguish
between two fish swimming close to-
gether, is about 0.5 s.
RESULTS
Calibration.
The counter was placed in the Ellidaár,
which is situated within the Reykjavík
City limits. This river has a mechanical
counter against which the electronic
counter was to be tested. During the 1975
season the counter was placed 400-500 m
below the weir. At this time the counter
did not include the printer, and a pen
recorder was used. It was impossible to
distinguish the direction of movement on
the record. The results are represented as
•veekÍy cumulative run (Fig. 5) and show
that the electronic counter consistently
registered a higher count at all times,
caused by some fish migrating
downstream after having first crossed the
mat upstream. After 7-8 weeks, when ac-
tive ascent was over and spawning fish
started to mill back and forth across the
mat, an ever-increasing discrepancy
arose.
Fig. 4. Decrease in counter sensitivity with incre-
asing distance of the fish from the electrode or with
decrease in length of the fish relative to the spacing
between the electrodes.
At the start of the 1976 season the
counter mat was moved up above the
weir, within 20 m of it. The printing unit
was in use throughout the summer. The
run was not monitored contionuously un-
til after August 10. After this time, until
the end of the angling season on Sep-
tember 9, the weir trap was closed 4 days a
week. While the weir was closed all the
fish were counted daily and most were
released above the weir. The electronic
count compared favorably with the
number of fish released above the weir on
these days (Table 1). However, during the
periods when the trap was open, the com-
parison was not as good between the two
counters. Comparison of the number of
fish in the trap during the days it was
closed, with the number shown on the
mechanical counter, indicated that the
mechanical counter was faulty.
Therefore, at the end of two seasons,
although favorable data had been ob-