FISHERY BULLETIN: VOL. 69. NO. 3 



Table 10. — Regressions of median bloctc densities and 

 dry weight concentrations on temperature and distance 

 from land.' 



^ N = number of sampling blocks; 

 = standard deviation about tne line; 

 * p = 0,05. 

 •• p = O.OI. 



= intercept; h = slope; 

 correlation coefficient. 



Table 11. — Average temperatures, median copepod den- 

 sities, and median dry weight concentrations for the 

 night periods of each cruise. 



Cruise 

 dots 



Temperature 



Small 

 copepods 



Large 

 copepods 



Dry 



weight 



water temperature was approximately the same 

 in both years. The copepod values and dry 

 weight consequently show a strong inverse re- 

 lation with temperature and date for 1961 but 

 not for 1962. 



The chaetognaths show inverse trends with 

 distance from the mainland and from nearest 

 land, but the former is the more significant of 

 the two. The relationship with distance to 

 nearest land includes many of the distance 

 measurements to the mainland, of course, and it 

 is possible that these are largely responsible for 

 the significant relationship with distance to 

 nearest land. The geographical distribution of 

 all chaetognath block medians (Figure 9) shows 

 that distance from the mainland is the more 

 pertinent independent variable. Low densities 

 occurred at all distances beyond 7 miles, where- 

 as the highest densities did not occur farther off- 

 shore than 14 miles, with a single exception. It 

 can be .seen that density was far more variable 

 near the mainland than offshore. 



DISCUSSION 



The food potential of plankton for pelagic 

 fishes depends on the relation between the aver- 

 age density of some or all species groups over 

 an area and the rate at which the fishes can feed 

 on these species groups. Since median density 

 of the more common species groups varies widely 

 within the space of a few months for the area 

 surveyed in this study, it is in'obabie that food 

 potential of near-surface plankton off southern 

 California fluctuates appreciably within short 

 time intervals. However, there were marked 

 small-scale variations in the distributions of den- 

 sities associated with the general area levels, and 

 the general level would not be an appropriate 

 index of food potential if fishes tend to orient 

 to small-scale features of distribution. 



Although the association of range with median 

 for sampling blocks of 51.8 km^ demonstrates 

 that densities vary, sometimes widely, within the 

 blocks, the medians of blocks of this or some 

 similar size probably constitute a scale of suf- 

 ficient resolution for assessing the food potential 

 of plankton over a large area. Maximum values 

 within the blocks are not likely to be imjiressively 

 greater than the median. If the median of large 

 copepods is 175 /m^, the block can be expected 

 to have, on the average, a maximum density of 

 245 /m-' representing an area of 2.6 km- ( 1 square 

 mile) within the block. If the median of 

 eu]jhausiids is 90 m\ the block can be expected 

 to have a maximum density of 130/m^ for 

 2.6 km-. Blocks with such medians are usually 

 rare, and the medians, as well as the maximum 

 values, are likely to be considerably higher than 

 the densities in most blocks in the area. The 

 medians would slightly underestimate the food 

 potential of such blocks only if plankton feeding 

 fishes tend to orient to the highest densities with- 

 in the space of 20 square miles. 



The distributions of .sampling block medians, 

 which were skewed unless general area level 

 was ver.v high, suggest that even under the 

 poorest general conditions relatively high den- 

 sities of organisms are likely to exist in some 

 small portion of the survey area. Small cope- 

 l)()ds, for example, showed a few occui'rences 

 of blocks with medians above 6000/m'', and one 



694 



