Theilacker and Porter: Condition of Theragra chalcogramma in the Gulf of Alaska 



337 



larvae grew (Fig. 4). Midgut cell heights of 

 fed larvae were significantly larger ( ANOVA, 

 P=0.006, df=l) than those of starved larvae 

 of equal length (4.50-5.49 mm SL), and the 

 same treatment was not significantly differ- 

 ent between years (ANOVA, P=0.533, df=2; 

 Table 2). Moreover, the difference in midgut 

 cell heights was attained after food was with- 

 held for 1 day (t-test; P<0.01, df=8; mean fed 

 midgut cell height after 1 d of feeding=12.96 

 ±1.31 p, n=5; mean starved midgut cell 

 height after 1 d of starvation = 9.12 ±0.99 

 pm, n=5; 1991 rearing data). 



Midgut cell height was not affected by a 

 net treatment of 7.5 min, the estimated time 

 a larva is handled during field collection. The 

 cell height of fish that were net-treated be- 

 fore being placed into preservative was not 

 significantly different from those placed di- 

 rectly into preservative U-test; P=0.12; 

 df=30). Additionally, midgut cell height was 

 not influenced by preservative type when 

 Bouin's solution was compared with Z-Fix (t- 

 test;P=0.17;df=80). 



Fixative effects 



Larval walleye pollock shrank an average of 

 4.7% in Z-Fix, 13% in Bouin's, and 8% in 5% 

 formalin. There was no measurable shrink- 

 age in 95% ethanol. For live larvae placed 

 directly in Z-Fix, shrinkage was a constant 

 proportion of size (Table 3). Laboratory 

 shrinkage in Bouin's and 5% formalin was 

 not a constant proportion of size; in each case 

 the y-intercept of the regression differed sig- 

 nificantly from zero (P=0.04 for Bouin's, and 

 P=0.001 for 5% formalin). Smaller larvae 

 shrank proportionately more than larger lar- 

 vae in both preservatives (Fig. 5A; Table 3). 

 For example, in using the regression equa- 

 tions in Table 3, the decrease in size of 5- 

 mm-SL larvae preserved in Bouin's was 15% and in 

 5% formalin was 9%, whereas the decrease in size of 

 7-mm-SL larvae in Bouin's was 12% and in formalin 

 was 7%. However, there was no observed shrinkage 

 of larvae preserved in 95% ethanol. In fact, the aver- 

 age larva increased slightly in size, about 1% for 5- 

 mm-SL larvae and 3% for 7-mm-SL larvae (Table 3). 



Net-shrinkage effects 



Larvae shrank an average of 7% after a 5-min net 

 treatment, 14% after 10 min, 17% after 15 min, and 

 22% after 20 min; initial sizes ranged from 4.48 to 



3.0 3.6 40 4.5 5.0 5.5 6.0 6.5 7.0 7.5 80 8.5 

 Laboratory - preserved standard length (mm) 



Figure 4 



Midgut cell height of laboratory-reared prefeeding, fed, and starved 

 larval walleye pollock, Theragra chalcogramma, related to labora- 

 tory-preserved standard length in Bouin's fixative. 



9.90 mm SL (Table 3), and sizes after net treatment 

 ranged from 3.76 to 9.48 mm SL. Length after net 

 treatment was linearly related to initial standard 

 length (net time=0) for each treatment period (Fig. 

 5B; Table 3). Analysis of covariance (ANCOVA) 

 showed that it was possible to interpolate a 7. 5-min 

 regression (needed for field procedures; see Field 

 Collections in Methods section; Table 3) between the 

 5- and 10-min net-shrinkage regressions; the slopes 

 of the two regressions were not significantly differ- 

 ent (P=0.785) and the lines were not coincident 

 (P<0.01). 



