ROTHLISBERG and MILLER: LARVAL PANDALUS JORDANI OFF OREGON 



shrimp were estimated from commercial landings 1.5 

 yr later (November) when they were fully recruited to 

 the fishing gear. Figure 14 shows the fluctuations in 

 these data from 1961 to 1980. Instantaneous rate of 

 total mortality, i, was calculated from the data ac- 

 cording to formulae of Ricker (1958). Results are 

 shown in Table 2. 



Year Class 



FIGURE 14. — Number of ova and numberofage I shrimp recruited to 

 the commercial fishery 1.5 yr later, by year class (Robinson 

 unpubl. data). 



Separation of the planktonic larval mortality from 

 total mortality required an estimate of juvenile mor- 

 tality. Gotshall (1969,1972), using an analysis of 

 stomach contents of Pacific hake, Merluccius produc- 

 tus, which has been shown to prey on P. jordani 

 without respect to age, has calculated that juvenile 

 mortality rate, i JUV , is 1.50/yr. Proportioning total in- 

 stantaneous mortality within the 18 mo from hatch- 

 ing to first harvest was done according to: 



i = 



0.274 (i larva ,)+ 1.226 (t juv ) 

 1.5 



where i = \n(N,/N )/t, with t = 1.5 yr 

 i uv = 1.50 (Gotshall 1969, 1972) 

 f' lan , al = instantaneous rate of larval mortality. 



Solving , 



_ 1.5(0 ~ 1.226(i iuv ) 



0.274 



Computed values of i Iarval , along with overall larval 

 survival (e~ a ), from commercial catch sample es- 

 timates for 1961 to 1980 are shown in Table 2 and 

 compared with our results for 1971 and 1972. Both 

 sets of estimates indicated that 1971 was a very poor 

 year for larval survival, while 1972 was slightly 

 above average. 



Factors Affecting Larval Survival 



Comparison of onshore-offshore distribution pat- 

 terns and wind patterns between 1971 and 1972 sug- 

 gested that Bakun's upwelling index (1973, footnote 

 4) might be a useful indicator of habitat quality for 

 survival of larval P. jordani. A study of possible cor- 

 relations was originally made with data from 1961 to 

 1973. Fortnightly, monthly, and quarterly indices 

 were calculated and regressed against the apparent 

 overall larval-juvenile survival (e~ a X 10 4 ). The July 

 index was the most highly correlated for a single month 

 {R = 0.599), while the average for the June through 

 August period gave the highest correlation found (R 

 = 0.749). For those months the equation 



Larval survival = e " X 10 4 = 0.73 

 (upwelling index) 



8.31 



accounted for 56.27c of variability in larval survival. A 

 new equation based on data from 1961 through 

 1980, 



Larval survival = e "X 10 4 = 0.889 

 (upwelling index) - 



9.652, 



accounts for 54.9% of the 20 yr variability in larval 

 survival. The slope is significantly different from zero 

 (t ls = 4.81, P<0.001). The pointfor 1967 contributes 

 strongly to the strength of the relationship, but the 

 regression is significant without it {t 17 = 3.35, 

 P<0.01). The relationship between larval survival 

 and upwelling index is shown in Figure 15. 

 This relationship of survival to upwelling strength 



100 



80 



O 

 x 



Ji 60 



"5 

 > 



> 



J) 40 



"5 

 > 



o 



20 



20 40 60 80 100 



June- August Upwelling Index 



120 



FIGURE 15. — Relationship between average upwelling in June 

 through August 1961 to 1980 and estimated larval survival (e~" X 

 10 4 ) 1961 to 1980, with fitted regression line. 



469 



