FISHERY BULLETIN: VOL. 69, NO. 



Formalin (in 50% seawater). Within 1 week of 

 sampling they were rinsed in distilled water 

 and their yolk dissected off, separated larvae and 

 yolk masses being placed on microscope slides. 

 The samples on slides were dried to constant 

 weight at 60° C and weighed to the nearest 

 microgram on an electrobalance. In agreement 

 with the results of Blaxter and Hempel (1966), 

 no effect of Formalin on the dry weight of lar- 

 vae was found, nor was there a significant effect 

 of Formalin on the dry weight of yolk masses 

 (this was tested in a previous experiment by 

 comparing dry weights of yolk dissected from 

 preserved larvae with yolk dissected from fro- 

 zen larvae or collected in preweighed capillary 

 tubes) . 



Larvae sampled from the supply containers 

 at 3-day intervals, and those collected from the 

 delayed-feeding series on day 20, were analyzed 

 for their ash, carbon, hydrogen, and nitrogen 

 content. Percent ash was determined by weigh- 

 ing separately three randomly chosen larvae 

 from each sample before and after combustion 

 at 500° to 520° C. In the case of larvae fed 

 from day 16 and sampled on day 20, only one 

 larva was available for the ash determination. 

 During combustion, larvae were held on tarred 

 pieces of precombusted aluminum foil, and 

 weighings were made on an electrobalance. The 

 remaining larvae in each sample were ground 

 into fine particles with an agate mortar and 

 pestle, and two aliquots of this material from 

 each sample were analyzed for carbon, hydrogen, 

 and nitrogen content with a Model 185 carbon- 

 hydrogen-nitrogen analyzer.' The number of 

 replicates was limited by the amount of material 

 available, but variation between replicate deter- 

 minations was small, and means calculated from 

 the replicates were accepted as the ash, C, H, and 

 N values for the sample. This appi'oach to the 

 chemical analysis of larvae was chosen because 

 it allowed determination of C,/'N ratios, estima- 

 tion of protein and fat content, and calculation 

 of caloric content (.see Results section). 



Larvae which had been used in the feeding 

 incidence experiments described above were pre- 

 served in 3'f Formalin and later cleared in KOH 

 and stained with Alizarin Red-S, the standard 



stain for bone, to allow comparison of ossifica- 

 tion in fed and unfed larvae. 



RESULTS 



BEHAVIOR 



Newly hatched grunion larvae have functional 

 eyes and jaws and are extremely active (Thomp- 

 son and Thompson, 1919; David, 1939). Grun- 

 ion larvae which received food in the present 

 experiment remained very active as they grew, 

 and some schooling behavior was noted as early 

 as day 6. Of more immediate concern was the 

 behavior of starved larvae. On day 7 it was 

 noted that unfed larvae were much easier to 

 catch with a pipette than fed larvae. As the 

 period of starvation lengthened, larval activity 

 declined and the number of larvae remaining 

 quiescent on the bottom Increased. Near the end 

 of the experiment, no starved larvae were swim- 

 ming freely above the bottom, and their activity 

 consisted in occasional erratic movements, fol- 

 lowed by long quiescent periods. 



SURVIVAL 



Figure 1 shows the survival to day 20 of 

 larvae which were fed at various times after 



DAY OF 

 FIRST FEEDING 



AGE (doysl 



Figure 1. — Survival curves for larvae with different 

 times of initial feeding, at 18° C. The number at the 

 end of each curve indicates the day of initial feeding. 

 The control group was given no food during the exper- 

 iment. 



' Hewlett Packard Corporation, Palo Alto, Calif. 



■416 



