Íslenskar landbúnaðarrannsóknir - 01.09.1978, Síða 96

94 ÍSLENZKAR LANDBÚNAÐARRANNSÓKNIR
the relationship of weight at return to the
time of release as well as the relationship
of time of return to the time of release.
Releases above and below the lagoon.
Fig. 10 shows the survival for smolts re-
leased conventionally from a rearing pond
above the lagoon and for those trans-
ported to a man-made channel close to the
sea. The results indicate that the conven-
tional releases are superior to the other
type. This is true for all smolt types. The
only smolts showing some reasonable sur-
vival when released below the lagoon were
1-year smolts which were kept in warm
water in order to enhance their smoltifi-
cation. Why this group was superior to the
others cannot be explained in this context.
The outcome of this experiment is
slightly contradictory since releases in a
similar manner were reported in the 1973
tagging experiment (Isaksson, 1976) and
gave 10—15% return. This phase of the
experiment, therefore, is inconclusive and
should be repeated.
Release time
Survival.
In discussing survival for different release
times, we will be confined to the groups
released above the lagoon, since only they
give resonable survival. The results are
presented in Fig. 10.
If we look at the 1 -year-normal and 2-
year-outdoor smolts which are the only
smolts with releases on all three dates, it is
clear that the June releases are best, fol-
lowed by May releases, but that the April
releases are very poor. These results show
that the traditional release of smolts in late
May as performed in the past is not neces-
sarily the right release time. The release
time should probably be keyed more to
climatic conditions in the spring and the
smolts should be released only when they
aré ready to migrate. The reason for the
low survival in May compared to June is
probably the fact that the fish released in
May were not ready for migration; they
probably stayed in the watershed for a
month or more without food, being ex-
posed to predation by birds and fish. The
June-released smolts, on the other hand,
were presumably ready to migrate when
released and probably did so in a few days.
As previously discussed, temperature
plays a key role in smolt migration; thus, a
cold spring could delay migration consid-
érably. The spring of 1975 was unusually
cold, as evidenced by the abnormally late
migration of natural smolts from Ellidaár
river that spring (Isaksson, Rasch, and
Poe 1978). The peak of the smolt migra-
tion probably occured around June 15,
but previous studies (Gudjónsson, 1954)
have indicated that peak smolt migration
in part of Iceland takes place around June
1.
April releases are clearly not suitable
for the smolts concerned. It might be pos-
sible, however, by the heating of rearing
water to produce smolts which might be
ready to emigrate in April. One such
group, 1-year early, was used in the 1975
experiment, but due to lack of smolts was
only released below the lagoon in April.
This group was actually the only group
that did have any significant survival
when released in April, which might indi-
cate the potential of research in that area.
The smolts in Fig. 10 are listed from left
to right in the order of delayed smoltifica-
tion; thus, the 1-year-early smolts had 700