A total of 247272 microtagged Atlantic salmon (Salmo salar L.) smelts
of four different year-classes (offsprings of 194 sires and 512 dams o
f nine different salmon stocks) were released to the sea from five rel
ease sites in Iceland. All returning grilse (one year at sea) and two-
sea-winter salmon (two years at sea) were slaughtered and their indivi
dual body weight and sex recorded. Strayers were included in the retur
ns of the site from where they were released. The return rates of gril
se varied from 0.41 to 3.85% and the return rates of two-sea-winter sa
lmon varied from 0.09 to 0.81%, The mean body weight of grilse varied
from 2.14 to 3.00 kg and the mean body weight of two-sea-winter salmon
varied from 5.09 to 6.32 kg, Mean total return biomass varied from 35
.3 to 101.6 kg. The estimated across year-class heritability for retur
n rate (on the underlying liability scale) was 0.12 for grilse, 0.04 f
or two-sea-winter salmon and 0.08 for total return rate; for body weig
ht, it was 0.36 +/- 0.11 for grilse and 0.00 +/- 0.15 for two-sea-wint
er salmon. The genetic correlations of grilse's return rate with total
return rate (0.98 +/- 0.01) and with total return biomass (0.90 +/- 0
.04) were very high. The genetic correlations between the other traits
recorded were all not significantly different from zero. The return r
ate of grilse is therefore the most important economic trait for sea-r
anched Atlantic salmon in Iceland. Averaged over the four year-classes
, the sire-by-release-site interaction effect explained 0.55% of the t
otal variation in grilse's return rate, while the sire component expla
ined 3.13% and the dam component 3.45%. (C) 1997 Elsevier Science B.V.