towards the siphon. The outflow tube of the siphon 

 recurved after passing through the top stopper of 

 the cyhnder and terminated in a small perforated 

 cylinder above the water surface. Inside the small 

 cylinder the outflow tube opened into a bag of 

 nylon netting with 0.024 mm mesh openings, 

 which removed food organisms and allowed water 

 to return to the rearing container by dropping onto 

 a small glass plate at the water surface. 



Operating at a flow rate of 3 1/h in containers of 

 12.6 1 capacity, the filter reduced Brachionus pop- 

 ulations ranging from 40 to 100 /ml at the start to 

 7 /ml in 1 d, 3/mi in 2 d, and considerably < 1/ml in 

 3 d. It reduced populations of Gymnodinium at 

 about the same rate. At the filtering rate of 3 1/hr, 

 larvae close to the submerged barrier cylinder did 

 not appear to be disturbed. 



Results 



Laboratory material showed good association 

 between food treatment and PAS glycogen rating 

 over the entire length range involved, 3 to 15 mm 

 SL. The larvae were divided into two groups, those 

 smaller than the median (6.4 mm SL) and those 

 larger than the median. For both size groups, as 

 for all larvae pooled, the majority of High ratings 

 occurred in the "fed" category and the majority of 

 Low ratings occurred in the "starved" category 

 (Table 1). It is noteworthy that only 4 larvae 

 among the 45 from starvation treatments received 

 High ratings. Three of these were from the oldest 

 age group, 26 d (starved 3 d), and were in fact three 

 of the four largest larvae in the entire starved 

 contingent. Examples of stained liver sections 

 representing High, Medium, and Low glycogen 

 ratings are shown in Figure 1. 



Larvae from the laboratory trials were fixed 

 in Bouin's solution, as described earlier, to be 

 comparable to the Bouin-fixed sea samples, but a 

 few other fixatives were tested as a matter of 

 perspective. Gendre's solution (Preece 1965) gave 

 somewhat better results than Bouin's, but 10% 



Table l. — Distribution of High. Medium, and Low glycogen 

 ratings, based on PAS staining intensity in livers of northern 

 anchovy larvae fed or starved in the laboratory. 



alcoholic Formalin,^ which is theoretically a 

 better fixative for preserving glycogen (Davenport 

 1960; Preece 1965), gave the best results. Some 

 specimens fixed in alcoholic Formalin from fed 

 containers had most hepatocytes solidly filled 

 with deep red color ( Figure 1 ) . Presumably more of 

 the glycogen present in the livers at termination 

 was retained with this fixation. However, integ- 

 rity of cells and tissues of the larvae was not well 

 preserved in alocholic Formalin. 



High, Medium, and Low ratings for the ocean- 

 caught larvae (Figure 2) reflect essentially the 

 same levels of staining intensity as for laboratory 

 material. Ocean samples show an appreciably 

 higher proportion of Medium ratings than the 

 laboratory material, but nevertheless exhibit 

 some association between glycogen rating and 

 histological characterization of the samples from 

 which the larvae were drawn (Table 2). However, 

 the association occurs only among the smaller 

 larvae, as indicated by comparing the distribu- 

 tions for larvae smaller and larger than the 

 median, 6.9 mm SL. The smaller larvae show a 

 relatively good proportion of High ratings for 

 robust samples and of Low ratings for emaciated 

 samples. The larger larvae do not exhibit this 

 association. 



Significance of the various distributions de- 

 scribed above is indicated by X^ values (Table 3) 

 calculated for each of the six pairs of columns 

 (treated as 2 x 3 contingency tables) from Table 1 

 and 2. All three X'^ values for the laboratory 

 material clearly reject the null hypothesis, i.e., 

 that there is no association between the two 

 classifications, food treatment and liver glycogen 

 rating. For the ocean material the null hypothesis 

 is rejected only for small larvae, indicating 



' Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service. NOAA. 



Table 2.— Distribution of High, Medium, and Low glycogen 

 ratings, based on PAS staining intensity in livers of northern 

 anchovy larvae from ocean samples showing either generally 

 robust condition or generally emaciated condition. Condition of 

 samples is based on a previous histological study of other 

 specimens from the same collection. 



808 



