Íslenskar landbúnaðarrannsóknir - 01.09.1978, Page 81

AN EVALUATION OF TWO TAGGING METHODS 79
ponds before the onset of their second
winter. The outdoor ponds are square,
Danish-type gravel-bottomed ponds. The
smolts are subjected to natural tempera-
tures all winter in the outdoor ponds.
2-year-photoperiod.
Instead of being in outdoor ponds, these
smolts are put in concrete raceways in a
rearing house with a transparent roof
which ensures natural photope'riod. These
fish are in natural temperatures through-
out the winter.
1-year-norrnal.
Like the previous group, these fish are
kept in concrete raceways in a rearing
house with a transparent roof. These fish
were kept in elevated temperatures
(9-10°C) starting in early March.
1-year-early.
With respect to photoperiod, these fish get
the same treatment as the previous group
but they get elevated temperatures
(9-10°C) starting in early February which
should enhance the moltification process
if temperature plays an important role.
1- year-late.
These smolts are the same as the above
except for temperature which was natural
throughout the winter- a condition which
might delay smoltification compared with
the other groups.
In addition to the above differences, the
fish usually have some differences in fin
condition. The fish of the 1-year and 2-
year photoperiods tend to have less ero-
sion of pectoral and ventral fins than the
2- year outdoor smolts. This is especially
true in the smaller-size categories.
ANALYTICAL TECHNIQUES
When testing for differences in survival or
mean weights between groups, one-way or
two-way analysis of variance was used,
depending on the experimental design.
When the F ratio was found to be signific-
ant at the 0.05 level and there was no
interaction, a least significant diíference
test was made to test for pairwise diífer-
ences (e. g., Snedecor and Cochran 1967).
When testing for diíferences in time of
return, replicates were pooled and
X1 2techniques were used, again at the 0.05
level.
The confidence intervals computed for
survival rates and tag loss rates assume a
binomial distribution. The confidence
intervals for mean weights assume a nor-
mal distribution.
COMPARISON OF MICROTAGS
AND CARLIN TAGS FOR MEASUR-
ING ATLANTIC SALMON SMOLT
SURVIVALS
The experiments
These experiments were conducted dur-
ing the 1974 and 1975 tagging seasons.
The primary purpose was to calibrate the
Carlin tag by comparing the survival of
fish carrying this tag to the survival of
microtagged fish in the same size categ-
ory, and to use microtags to check on
dangler-tag loss. All the microtagging in
1974 was done by one of the authors,
Bergman, who was experienced in tagging
Pacific salmon. He prepared head molds
for the various sizes of smolts used (fig. 2),
and was assisted by Icelandic personnel
who did all the Carlin tagging and adipose
clipping in preparation for the mic-