eyed larvae. We therefore did not use UV irradia- 

 tion routinely. If UV irradiation is used to de- 

 grade retinol, the exposure must be kept short (1 

 or 2 minutes), or else another compound fluoresc- 

 ing at these wavelengths appears (Fig. 3), giving 

 the spurious impression that UV did not affect 

 retinol; this may explain the negative results of 

 Csallany and Ayaz (1976). 



To evaluate the effect of preservation in forma- 

 lin, we froze subsamples from stocks of larvae, 

 and preserved other subsamples in 10% formalin 

 in a glass container, using formalin from a glass 

 reagent bottle. We analyzed frozen and preserved 



80,- 



4 6 



MINUTES 



Figure 3. — Time course of fluorescence resulting from UV 

 irradiation of duplicate batches (two symbols) of retinol. 



larvae after 3 months, and after an additional 

 year we analyzed more preserved larvae. 



Nicol (1987) recently demonstrated that forma- 

 lin preservation significantly increases the fluo- 

 rescence of lipofuscin extracted in chloroform- 

 methanol, and our results show similar analytical 

 problems (Table 1). There was significant en- 

 hancement in fluorescence in the larvae pre- 

 served for over a year, when compared with those 

 preserved for 3 months. Quinine sulfate stan- 

 dards run at the same time as the samples for 

 different periods were very similar (Fig. 4), indi- 

 cating real changes with time of preservation. 

 Thus, this method will not permit using historical 

 collections of formalin-preserved animals to de- 

 termine their relative physiological states until 

 more is known about the time course and nature 

 of the effect of formalin on the extract of pre- 

 served tissue. 



Larval Growth Experiments 



Seabass larvae were obtained from Hubbs Re- 

 search Institute and halibut larvae from the Los 

 Angeles County Natural History Museum; larval 

 grunion were obtained by stripping adults and 

 bringing fertilized eggs into the laboratory for 

 hatching in an aerated, 10 L container of sea- 

 water at room temperature. 



Larval grunion were reared in 40 L Nalgene 

 tubs with spigots at the bottom (which facilitated 

 emptying and cleaning), initially stocked with 

 100—120 newly hatched grunion per tub, with 

 replicated "high" and "low" food concentrations. 

 Larvae were fed newly hatched Artemia nauplii 

 and the rotifer, Brachionus , which was cultured 

 on Dunaliella tertiolecta in 100 L, lighted 



Table 1 . — Lipofuscin (as fluorescence units, FU) per animal and per unit dry weight (DW) of frozen vs. formalin- 

 preserved larval fish. N = number of larvae per analysis. 



410 



