FISHERY BULLETIN: VOL. 72, NO, 1 



MORTALITY 



There are several limitations to the data used 

 for estimating natural mortality: (1) Sampling 

 was conducted during a period of apparent 

 migration which caused the stock along the 

 Pacific coast to become stratified in size and 

 age composition. (2) sampling was limited 

 and not conducted throughout the stock's 

 entire geographical range, and (3) age frequency 

 data indicate possible variations in annual 

 recruitment and/or survival rate. Thus, one 

 or more basic assumptions underlying tradition- 

 al mortality models are violated to some degree. 

 Realizing the above limitations and considering 

 this an initial study of the adult stock, I have 

 generated a- range of estimates using several 

 independent techniques. 



A catch curve analysis (Robson and Chapman. 

 1961) was applied to the 1970 data since equal 

 sampling occurred off Washington and Oregon, 

 the only areas sampled, during that period. 



= 997 



Annual survival rate estimate: 



S 



Var (.s) 



SE (.s) 



95% CI (.s) 



.2876 

 0.00017 

 0.013 



.2876+ 2(0.013) 



= (.2616, .3136) 



When the above data were converted to a total 

 annual in.stantaneous mortality rate, Z = 1.25 

 and 95% CI, Z = (1.16, 1.34). 



The raw age data were also converted to 

 natural log form and treated by simple linear 

 regression. The result was a significant linear 

 relationship with Z = 1.67 and 95% CI, Z = 

 (1.41, 1.93). 



Beverton and Holt's (1956) formula using 

 length frequency data was also applied to the 

 1970 data. While this technique was designed 

 primarily for exploited pojjulations, its use 



generated another independent estimate and 

 enabled the use of substantial numbers of fish 

 which could not be aged. 



The Beverton and Holt formula Z = 



KiL^-L) 



where L is the average length of 



(L - L,) 



the fish in the catch that are as large as, or 

 larger than, the first fully recruited length 

 Lf, estimated Z = 1.41 when L^ = 342 mm, 

 K = 0.41, L = 248 mm, and L^ = 220 mm. 

 Survival rates were generated for each of 

 the four area-year categories by subjecting 

 resijective sets of age frequency data to Jackson's 

 (1939) technique: 



S = 



N^^N,^ 



+ N„ 



A^i + A^2 + 



+ N-1 



The analysis yielded the following estimates: 

 Oregon, 1970; Z = 1.58, 

 Oregon, 1971; Z = 1.80, 

 Washington, 1970; Z = 2.20. 

 Washington-Vancouver Island, 1971; 

 Z = 1.27. 

 The seven individual estimates obtained 

 indicate a possible range of Z from 1.25 to 2.20 

 and an overall average Z of 1.60. 



AVAILABILITY OF FISHABLE 

 CONCENTRATIONS 



Fishable concentrations of saury (>% ton) 

 were usually located in waters of 15°-17°C 

 near areas of upwelling. Surface temperatures 

 strongly influence distribution and migration 

 patterns of western Pacific saury (Fukushima, 

 1956 and 1962) as they appear to in the eastern 

 Pacific (Ellis and Hughes, 1971). All studies 

 indicate sharp thermal fronts affect and often 

 dictate patterns of migration and areas where 

 temporary concentrations may form. 



Throughout the surveys, high density areas 

 capable of sustaining productive fishing opera- 

 tions were rarely encountered. Most encounters 

 were single schools (1-3 tons) or loose aggrega- 

 tions offish dispersed over large areas of surface 

 waters. The average probability of locating 

 at lea.st one fishable concentration during a 

 night's oi)eration (averaging 8 hr of searching- 

 effort and 70 miles of tracklines) was about 0.3. 

 The relative densities decreased slightly with 

 increasing latitudes, but large saury, which are 

 currently in greatest commercial demand, were 



128 



