In the field, contamination is most likely to 

 occur between consecutive samples that are 

 taken without interruption. In a discrete vertical 

 series, where samples are taken from depths 

 that are distinctly separated, contamination 

 between depths can be reduced by flushing the 

 system with water from each new depth before 

 collecting. The cumulative percentages (table 

 1, column 6) offer a guide for judging the 

 presample flushing volume that is needed in 

 the field. They indicate the probability of 

 passing a shrimp through the system in rela- 

 tion to the volume of water pumped after its 

 introduction. If we assume that planktonic 

 organisms become dispersed as the brine 



shrimp were, the cumulative percentages offer 

 conservative estimates of the portion of con- 

 taminating organisms that would be cleared 

 from the hose with different flushing volumes. 

 For example, a flushing volume of 700 liters 

 in the field should remove at least 91 percent 

 of the organisms that remain in the hose from 

 the previous sampling depth. 



Escapement Through Straining Apparatus 



After the organisms pass through the water- 

 meter, they enter a plastic funnel which has 

 screened windows and a cylindrical cod end at 

 the bottom (fig. 3). The screening is stainless 



Figure 3. — A filter funnel showing the screened windows, cod end, and rubber cap. 



