MURPHY and CLUTTER: PLANKTON PURSE SEINE 



Figure 1. — Underwater view of the plankton purse seine 

 just as pursing was initiated. Photograph by Robert R. 

 Harvey. 



CALIBRATION OF SEINE 



Although, as will be seen later, it is not ab- 

 solutely necessary for some analyses to know ex- 

 actly how much water was sampled by the seine, 

 it is desirable to have a reasonably accurate es- 

 timate. The size of the net and idealized geom- 

 etry of a perfect set suggest the net could sur- 

 round about 500 m^ of water. 



Actual sets are not perfect, and an attempt 

 was made to standardize for obvious imperfec- 



tions by noting visually estimated percent effi- 

 ciency during each set. These ranged from 60 

 to 100% with six from 60 to 65%, nine from 80 

 to 85%, and the balance (40) from 90 to 100%. 

 All numerical data generated by the samples 

 were adjusted to 100% efficiency. 



Since metered plankton net samples were 

 available from presumably the same water as the 

 purse seine, it should have been possible, ideally, 

 to calibrate each set by comparing the catches 

 of some nonmotile component of the plankton, 

 e.g., fish eggs. This was not possible because 

 such items were not consistently present and, 

 more importantly, because high amplitude, short- 

 period patchiness was obviously generating high 

 variability, even between such closely spaced 

 samples. 



The final decision was to use total sample vol- 

 ume as a measure of the water strained by the 

 seine, i.e., the volume of water strained by the 

 seine was estimated by comparing the total vol- 

 ume of wet plankton with the total volume of 

 wet plankton taken by the metered plankton net. 

 This was based on the hindsight fact that the 

 volume of non-escapers, e.g., ctenophores, was 

 very large compared with the volume of escapers, 

 e.g., fish larvae, and therefore, the total volume 

 was essentially independent of escapers. Several 

 analytical techniques, e.g., regression, were con- 

 sidered and applied. These yielded estimates of 

 the purse seine volume at 100% efficiency from 

 305 m^ to 441 m\ Finally adopted was a simple 

 comparison of the geometric means (because the 

 volumes were logarithmically distributed) of the 

 two series after raising the meter net catches 

 to 300 m^ of water strained, and the purse seine 

 catches to 100% efficiency. This yielded a value 

 of 356 m^ of water strained for the purse seine 

 at 100% efficiency. The procedure was: geo- 

 metric mean of 54 purse seine catches divided 

 by the geometric mean of 54 300-m^ plankton 

 net catches multiplied by 300 equals mean purse 

 seine volume of 356. This value, of course, re- 

 mains an approximation. Any error has no ef- 

 fect on the analytical portions of this paper as 

 the slopes of the length frequency curves are of 

 primary concern. As indicated earlier, the pri- 

 mary purpose of the adjustments is to deal with 

 reasonably real world numbers. 



791 



