We reviewed the biological and oceanographic data used to help decide
on the siting of salmon farms in Canada (British Columbia and New Brun
swick), Norway, Scotland, Ireland, Iceland, United States (Washington
and Maine). The LENKA program in Norway and a ranking technique based
on biophysical features in British Columbia are examples of semi-quant
itative approaches. A zoning system partially based on detailed data o
n fjord oceanography has been initiated in Scotland. Guidelines to pro
tect critical fish habitat, especially for wild salmon, have been deve
loped in each of the jurisdictions. Models to help predict sedimentati
on, hypernutrification, and dissolved oxygen levels have been used in
Norway and Washington. Siting criteria need to be supplemented with th
ese models, which consider cumulative effects and predict the carrying
capacity of key fjord areas where farms are likely to cluster. Coordi
nated and comprehensive research programs are needed which consider sa
lmon production in the context of the oceanographic regimes within fjo
rds. For site-specific evaluations and as communication methods, refer
ral systems and guidelines are relied on extensively in site evaluatio
ns in the various jurisdictions.