Fig. 1. A representation of the natural process that could determine the total number of salmon that survive in the ocean and return to spawn. In the upper panel the abundance of juveniles is greatly reduced immediately after juveniles leave fresh water (FW). By mid-July in salt water (SW) predation mortality after this period is relatively minor and there is little change in trend in the winter. In the lower panel, predation mortality accounts for about one half the loss in abundance by mid-July. There is minimal mortality until about mid-October when a second growth-based mortality occurs. The year-class size is virtually determined during the winter. Some mortality would occur after the first marine winter, but is small relative to the earlier mortalities. It is the second, growth-based mortality that links climate and salmon production, naturally.

Fig. 2. Location of the Strait of Georgia study area showing the track lines used in the surveys.

Fig. 3. Examples of the otolith weight, fish length relationship in 1999 between (a) September and (b) November. The solid line is the calculated regression. The change shown in this relationship was typical for all three years.

Fig. 4. The otolith weight frequencies in (a) September 1997, (b) November 1997, (c) September 1998, (d) November 1998 and (e) September 1999.

Fig. 5. The relationship between plasma IGF-I and length for ocean age 0 coho sampled in the feeding studies in seawater net-pens September 30, 1999 (○) and from samples collected in September 1999 in the Strait of Georgia (□) and in Puget Sound ().

Table 1. Minimal juvenile salmon abundance in the late summer in the Strait of Georgia and the comparable smolt production and total returns

Table 2. Weights (g) of samples of otoliths in relation to mean weight (g), and length (mm) of coho collected in September and October–November 1997, 1998 and 1999