FISHERY BULLETIN: VOL. 84, NO. 4 



Table 4.— Microzooplankton: relative density, May 

 1978-April 1979. 



Taxon 



mean 

 + 1 SE 



n-1 



lated with zooplankton density (r = 0.27, P = 0.027). 

 All interpretation of the microzooplankton data was 

 done under the assumption that estimates of volume 

 filtered are accurate. 



Environmental Variables 



The mean surface water temperature during this 

 study was 15.2°C. The coldest reading was 8.0°C 

 at station 2 in January; the warmest was 22.5 °C at 

 station 1 in August (Fig. 2). Water temperature near 

 the bottom varied from 8° to 21.5°C (mean = 

 15.0°C). Mean temperature stratification, the dif- 

 ference between the surface and bottom tempera- 

 tures, was 0.2°C. Stratification was generally pres- 

 ent June through October, especially at station 5. 

 Mean stratification during these months was 0.5°C 

 (Fig. 2). During February and March 1979 the sur- 

 face temperature was lower, on average, than the 

 temperature near the bottom thus showing the influ- 

 ence of air temperature on the surface water tem- 

 perature. Surface salinity varied from 3 to 31°/oo 

 (mean = 23.6%o). Bottom salinity was 14-31%o 

 (mean = 24.8%o). The low readings for both sur- 

 face and bottom salinity occurred at station 5 dur- 

 ing March 1979. Surface salinity at station 1 was 

 usually low, showing the influence of freshwater in- 

 flow at the south end of the Bay (Fig. 2). Salinity 

 at station 6 was relatively high, showing the oceanic 

 influence at the Golden Gate. Surface salinity at 

 other stations reflected their relative positions be- 

 tween these two influences. The lowest surface 

 salinity was always at station 5 due to the Sacra- 

 mento River discharge. During March 1979, salin- 

 ity at stations 4 and 6 also showed the effects of high 

 freshwater discharge which lowered the salinity at 

 station 5 to 3%o. Salinity was slightly lowered this 

 month at station 3 in South Bay also. Surface salin- 

 ity followed a seasonal pattern; it was high from July 

 through January and low in the winter and spring 

 months. Relatively high salinity corresponded to 



high temperature July through October. Salinity 

 stratification was generally <2%o except at station 

 5 where the average stratification was 4.7°/oo (Fig. 



2). 



Surface salinity was negatively correlated with 

 salinity stratification, (r = -0.62, P < 0.001), and 

 positively correlated with Secchi depth (r = 0.39, 

 P = 0.001). Salinity stratification was negatively 

 correlated with Secchi depth (r = - 0.29, P = 0.012). 



Turbidity 



Light penetration was lowest at stations 1 and 5, 

 and highest at stations 6, 4, and 3 (Fig. 2). The mean 

 depth of light penetration during this study was 1.1 

 m with a range of 0.1-2.5 m. The data suggest a 

 weak seasonal trend with light transmission higher 

 in summer and lower in winter. The variable with 

 the strongest linear association with Secchi depth 

 was zooplankton density. Light penetration was in- 

 versely related to zooplankton density (r = -0.58). 



Relationships Among Varibles 



Northern anchovy egg abundances were positively 

 associated with surface temperature, temperature 

 stratification, and Secchi disk depth and negative- 

 ly correlated with zooplankton density (Table 5). 

 Eggs were positively associated with larvae but this 

 correlation was not significant at the 5% level (P 

 = 0.053). Larvae were positively correlated with 

 surface temperature and zooplankton density (Table 

 5). They were negatively correlated with Secchi 

 depth. Thus, eggs and larvae both were significantly 

 correlated with zooplankton and Secchi depth but 

 in opposite directions: eggs were associated with 

 clearer water and lower zooplankton density, lar- 

 vae with more turbid water and higher zooplankton 

 density. 



Stepwise Multiple Regression 



Surface temperature alone explained 65% of the 

 variability in egg density (r 2 = 0.651). The combi- 

 nation of microzooplankton density with surface 

 temperature explains an additional 1.5% of the vari- 

 ability of egg density. The addition of all other 

 variables only increased the amount of variability 

 explained to 68% (r 2 = 0.678). The predictive 

 regression model using the independent variables 

 whose addition to the model improved its prediction 

 by more than 1% is 



E = -2.20 + 0.317T - 0.502M 



886 



