SALMON AND SALMON FISHERIES OF SWIFTSURE BANK 
711 
for 1887 which states that the fishermen go out only as far as the lightship, 4 or 5 
miles from land. 
Table 1 gives the number of canneries operated annually from 1876 to 1934. 
For nearly 20 years the proportion of the canneries located at New Westminster 
declined, while the proportion near Steveston and Ladner continued to rise. The 
few remaining canneries were either at the river mouth, in the North Arm, or entirely 
outside the river proper. 
The canneries at Ladner reached their peak in 1885, when half the total number 
operating were located there, and have since declined steadily to a point of little con- 
sequence. Many ascribe much of this decline to the fact that the fish have entered 
the river through Canoe Pass in decreasing numbers since the driving of traps at 
Point Roberts. The decline may possibly be further ascribed to the silting up of 
Canoe Pass and the change in the main channel at Woodwards Slough, effected dur- 
ing the flood of 1 894, which made it difficult to reach most of the canneries with large 
boats. 
SEASONAL OCCURRENCE OF EACH SPECIES 
Seasonal occurrence is of prime importance in any fishery wherein more than one 
species is taken, as the intensity of fishing for a species is not governed by its abundance 
alone, but by a combination of factors, such as the relative abundance of the several 
species at any time during the season, as well as the relative prices. 
In determining the seasonal occurrence for sockeyes, data for 1,982,735 fish taken 
in 30,706 gill-net deliveries were used, covering 3 complete 4-year cycles, 1898-1909, 
inclusive. The occurrence shown in these early years was considerably different than 
that shown in the last three cycles, 1923-34. This difference is treated in the sock- 
eye section of this report (see page 754). 
The king salmon curve is derived from 102,123 fish taken in 26,193 deliveries 
over a 5-year period, 1929-33. 
For pink salmon 8 years are represented, all of the odd-numbered years from 
1915-33, except 1917 and 1921; the data totaling 597,774 fish in 15,581 deliveries. 
The coho curve is also based on 8 years’ data, 1904, 1905, and 1929-34, and repre- 
sent 155,957 fish in 22,117 deliveries. 
The chum-salmon curve represents only 3 years, 1932-34, but is quite repre- 
sentative of those particular years, comprising 263,703 fish from 10,608 deliveries. 
In analyzing these data the average catch per delivery for each 7-day period was 
computed for each year and then given equal weight in determining the average 
curve for all years (see table 5). 
Table 5 shows that the period over which one or more species can be taken in some 
measure of abundance extends from June 24 (week ending June 30) to November 17; 
21 weeks, or 147 days. As mentioned above, in earlier years the season was very 
much shorter, corresponding largely to the more abundant portion of the sockeye run. 
The sockeye and pink-salmon runs, which overlap to a slight extent, are both 
of short duration. Approximately 79 percent of the pinks are caught in 4 weeks, 
September 2-29, and 83 percent of the sockeyes are taken in the 5 weeks from July 
22-August 25. 
