252 



Fishery Bulletin 105(2) 



Determination of egg stage and average egg size 



We determined the stage of development of the eggs 

 (egg stage) and measured the egg diameter and oil 

 globule diameter of 30 fresh, randomly selected eggs 

 from each standard collection made at 2 h after fertil- 

 ization. We measured the eggs to the nearest 0.1 mm 

 with a dissecting microscope fitted with an ocular 

 micrometer. Eggs were classified developmentally 

 according to the terms used by Fritzsche (1978). Egg- 

 stage duration was estimated as the time period from 

 fertilization to 50% hatching (defined below). 



incubation of eggs 



We incubated the eggs in conical fiberglass tanks con- 

 taining 300 L of l-fim-filtered, UV-sterilized seawater. 

 Through mid-1997, the incubation tanks were exposed 

 to ambient air temperatures and indirect ambient 

 light. Beginning in mid-1997, the incubation tanks 

 were housed in a room with indirect fluorescent light- 

 ing and at a constant temperature. The incubation 

 tanks were in total darkness during the night, and 

 were never exposed to direct sunlight or overhead 

 fluorescent light at any time. We used tanks with both 

 flow-through and closed systems for incubation, but 

 we detected no notable differences in hatching suc- 

 cess between the two systems. We rinsed the eggs in 

 a 500-,um sieve before placing them in the incubation 

 tanks to eliminate any potentially harmful organisms 

 such as Benedenia trematodes or parasitic copepods 

 that might have transferred with the egg sample from 

 the broodstock tank. Beginning in January 1997, we 

 recorded the daily temperatures of the incubation tanks 

 approximately four times between initial stocking and 

 the time of hatching. 



Determination of time at hatching 



To determine the time at 50% hatching, we collected 

 a sample of eggs from the incubation tank at 15-min 

 intervals, beginning about 12 to 15 h (depending 

 on water temperature) after the estimated time 

 of spawning. When 50% of the eggs in the sample 

 were hatched, we recorded the time. Approximately 

 2 hours after 50% hatching, we made a final esti- 

 mate of the mean hatching rate. We took three 300- 

 mL samples from the incubation tank and counted 

 the total number of dead and live eggs and yolksac 

 larvae. Live unhatched eggs, which made up a small 

 portion (<5%) of the samples, were considered to be 

 almost ready to hatch because our experience has 

 shown that the majority of these eggs eventually 

 hatch. The estimated final percent hatching was 

 calculated as 



Final % hatching = {no. larvae + no. live eggs) I 



{no. larvae + no. live eggs + no. dead eggs + 



no. dead larvae) x 100. 



Collection of yolksac larvae 



After all the larvae were hatched (final hatching), we 

 made morphometric measurements on 20 randomly 

 selected yolksac-stage larvae from each daily cohort. 

 For each larva we measured total length (TL), notochord 

 length (NL), yolk length, yolk height, and oil globule 

 diameter to the nearest 0.1 mm. 



Developmental series 



Periodically we followed the development of a daily 

 cohort of eggs and larvae from fertilization to first- 

 feeding (normally a duration of 3.5 days). We initiated 

 a developmental series whenever the daily mean tem- 

 perature of the broodstock tank changed by at least 1°C. 

 Each developmental series entailed sampling eggs at 15- 

 min intervals for the first 4 hours after fertilization, at 

 1-h intervals for the next 6 h, and then at 2-h intervals 

 until the time at 50% hatching. After final hatching, 

 we sampled yolksac larvae at 6-h intervals until the 

 larvae were ready to feed. We considered a larva to be 

 at first-feeding stage when its retina was pigmented, 

 the alimentary tract was formed, and the mouth was 

 fully developed. We took morphometric measurements 

 on 20-30 fresh eggs and yolksac larvae, as described 

 previously. At first feeding, we measured 20 live larvae 

 for TL, NL, the height and length of any remaining yolk, 

 and the diameter of the oil globule. We measured mouth 

 width on freshly fixed (5% formalin) first-feeding larvae 

 because of the difficulty in obtaining accurate mouth 

 measurements on live specimens. 



Dry weights of eggs, yolksac larvae, 

 and first-feeding larvae 



For each developmental series examined, we obtained 

 fresh dry weights of 20-30 eggs, yolksac larvae, and 

 first-feeding larvae. We used 8-mm diameter aluminum 

 pans that were dried in an oven at 60°C for 24 h, des- 

 iccated for 24 h, and then individually weighed to the 

 nearest 0.1 pg on a microbalance. After measuring, we 

 rinsed the larvae and eggs multiple times with distilled 

 water to remove salts and particulate matter. We placed 

 an individual egg or larva in a preweighed aluminum 

 pan, dried it at 60°C for 48 h, desiccated it for 48 h, and 

 then weighed the specimen to the nearest 0.1 i.tg. 



Data analysis 



We analyzed spawning parameters and the character- 

 istics of eggs and early-stage larvae in relation to bio- 

 logical and physical data, including water temperature, 

 time of day, and lunar cycle, from October 1996 through 

 March 2000. We analyzed the relationships of daily 

 ration and female size with egg size and egg production. 

 We considered 20 kg as the minimum size for actively 

 spawning females, based on the size at first-spawning 

 from the genetic analysis (Niwa et al., 2003) and our 

 observations of courtship and spawning behaviors. We 



