Íslenskar landbúnaðarrannsóknir - 01.09.1976, Qupperneq 13

TAGGING EXPERIMENTS AT KOLLAFJÖRBUR 11
the finclipped smolts showed a high return-
rate of 7,5% almost twice as high as corre-
sponding tagged smolts. This shows that the
return-rate of untagged fish must be at least
twice that of tagged fish, and possibly greater
for fish in higher length groups since the
remrn-rate of unmarked fish seems to be 3—
4 times that of tagged fish in this experiment
(see next paragraph).
The total return of salmon to the Fish
Farm in 1973 was 1956 salmon. Of those
46 are kelts which is approximately 30%
remrn from last year’s tagging. Of the re-
maining 1910 fish there are ca. 1600 remrn-
ing as grilse. This is primarily the result of
a release in 1972 of approximately 14.000
viable two-year-old smolts marked and un-
marked. This amounts to a total remrn of
approximately 11%. Subtracting tagged fish
from the total return of grilse gives ca. 1200
fish returning unmarked. This is the return
from 6000 two-year-olds released amounting
to approximately 20% return. A conservative
estimate would be ca. 17%.
4. The Length at Tagging versus Returns.
These data were set up as in the previous
experiment using only returns of two-year-
old smolts and are shown in Table 6.
The table shows that there is an increase
in remrn rates up to approximately 19 cm
length. After that it tends to level off and
even decline after a size of 21 cm is reached.
These findings are in good agreement with
those shown by Carlin (1969) although his
results levelled off at a much higher remrn-
rate. These results further suggest that tag-
ging with Carlin tags is not practical unless
the fish are at least 14 cm in average fork
length.
A comparison of the results obtained by
Carlin (1969), Guðjónsson (1972) and
those from the present experiment is present-
ed in graph 1. It is apparent that the results
from 1971 and 1972 tagging experiments
are quite a bit lower than Carlin’s data and
much lower than those of Guðjónsson. More
recent data (1973 experiment) have provided
considerably higher remrn-rates in every
length-group with less pronounced increase
with length. This information will be dealt
with in a later report.
The decrease of remrns in smolts larger
than 21 cm is not easy to explain but it can
be speculated that this is due to early sexual
mamrity of the male smolt and a tendency
to linger on in freshwater. This was in fact
noticed at the Fish Farm in the fall of 1972.