THEILACKER and WATAXABE: NLTRITIOXAL STATUS OF NORTHERN ANCHOVY 



increase noted in the midgut index (midgut cell 

 height/SL) of the Bouin's-fixed specimens (Fig. 

 6c, d) indicates that the model is overcorrecting 

 for Bouin's. That is. the additional shrinkage 

 caused by fixation (shrinkage in addition to that 

 due to abrasion by the collecting net) is less 

 when Bouin's is used for the final fixative than 

 when formalin is used. This logic, however, is 

 hard to reconcile because Bouin's. made with 

 2Q9c formalin, is stronger than the 5'i formalin 

 solution we used. 



Starved fish may be shrinking more during the 



net treatment than their fed counterparts, fur- 

 ther complicating the interpretation of the 

 shrinkage adjustment. But we cannot be certain 

 that this occurred because we did not follow indi- 

 vidual larvae during the net treatment. Yet 

 there was a significant decrease in length of 

 starved larvae when the adjusted lengths (cor- 

 rected to "live length") were regressed and plot- 

 ted over time (Fig. 7b; Table 6). The decrease in 

 length was not evident for the fed larvae (Fig. 

 7a). This apparent difference in shrinkage rates 

 between fed and starved larvae needs to be 

 tested. 



E 

 E 



X 



o 



z 



UJ 



5.0 10.0 15.0 20.0 

 TIME (min) 



25.0 



Q 4.5 



CC 



< 



G 4.0 



< 



3.5 - 



3.0 



b 



3 • • 



2 • • 



^ 2 • 2 2 



1 ' 



5 ^"2— -— ____^ • 3 



6 5 2 — ____^ • • 



— • • • ~~ — ~~-~~_4 



4 2 — — — -_. 



•  



3 2 



2 



• 2 



5.0 10.0 15.0 20.0 

 TIME (mIn) 



25.0 



Figure 7. — a) Change in standard length over 

 time for fed Engranlis mordax larvae that 

 were fixed in formalin and Bouin's. bj Change 

 in standard length over time for starved 

 EngrauUs mordax larvae that were fixed in 

 formalin and Bouin's. 



Table 6. — Coefficients for regressions of midgut indexes (midgut cell heigtit adjusted standard 

 length (mgtit/SL)) and of adjusted standard length (SL) with elapsed time-in-net (t). 



467 



