256 



Fishery Bulletin 105(2) 



o 200 



CL 



150 



100 



i2 50 * 



period (r=0.15, df=564, P>0.10). How- 

 ever, over short time periods (<1 month), 

 the spawning females increased their 

 egg production in response to short- 

 term increases in daily ration. Over the 

 period of June 1997 through July 1999, 

 we varied the daily ration of food for 

 the broodstock fish from approximately 

 1.0% to 4.5% body wt/day. During eight 

 periods, we purposely increased daily 

 food ration from 9% to 33% over time 

 durations of 3 to 14 days. The rations 

 were usually increased in response to 

 greater food requirements, signaled 

 by increasing feeding activity, or in 

 attempts to effect increased egg pro- 

 duction. During all eight of these peri- 

 ods, the standardized egg production in 

 the tank increased from 30% to 234% 

 (Fig. 4). There was a time lag until peak 

 egg production occurred (indicated by 

 numbers in parentheses in Fig. 4) after 

 the initial increase in ration. Egg pro- 

 duction increased, and peaked from 4 to 

 21 (average of 12) days after the intro- 

 duction of increased rations. The percentage increase 

 in egg production tended to be greater with greater 

 increases in daily ration, but the relationship was not 

 significant (/•=0.36, df=7, P>0.10). The increases in egg 

 production occurred over a range of water temperatures 

 from 27.3° to 29.4°C; however, there was no clear asso- 

 ciation between water temperature and standardized 

 egg production. 



Spawning and photoperiod 



Given the tropical latitude of the Achotines Laboratory 

 {7°25'N), photoperiod was relatively constant during 

 the study. From October 1996 through March 2000, 

 day length varied by only 53 min, and changed by 

 only 3 to 12 min/month (U.S. Naval Observatory 

 Astronomical Applications Database'^). No strong 

 relationships between either frequency of spawning 

 or standardized egg production and photoperiod were 

 apparent. Cessations in spawning associated with 

 short-term decreases in tank water temperature (in 

 December 1997, October 1998, and August 1999) 

 occurred during periods of decreasing photoperiod, 

 while cessations in spawning due to water temper- 

 ature decreases below the apparent-minimum for 

 spawning (<24°C) (in March 1997, March 1999, and 

 February 2000) occurred during periods of increasing 

 photoperiod. 



10 15 20 25 



Percentage increase in daily ration 



30 



35 



Figure 4 



Increases in egg production by female yellowfin tuna (Thunruis albacares) 

 broodstock during eight periods of planned increases in daily ration. Each 

 period is represented by a plotted symbol. For each period, the value in 

 parenthesis is the elapsed number of days from the initial increase in 

 ration until the peak egg production was observed. 



Spawning and lunar phase 



Spawning occurred almost daily and showed no defini- 

 tive relationship with lunar phase on a daily or monthly 

 basis. Also, there was no significant correlation between 

 standardized egg production and lunar phase (r=0.10, 

 df=564, P>0.20). In each of the 26 months for which 

 standardized egg production was calculated (June 1997 

 through July 1999), the monthly maxima egg produc- 

 tion occurred equally during the first two quarters (new 

 moon to half-moon) and last two quarters (half-moon to 

 full moon) of each lunar cycle. However, over the 2-year 

 period, the maximum egg production from individual 

 spawns occurred most often at greater illumination 

 phases of the moon. Of the 20 spawns with the greatest 

 standardized egg production, 90% occurred during the 

 third or fourth quarters (i.e., half-moon to full-moon), 

 although only one of these 20 spawns occurred directly 

 on a full moon. 



Egg-stage duration and egg size 



Egg-stage duration increased with larger egg size, 

 although the predictive power of the linear regression 

 was low (r'-=0.09, df=837, P<0.001). Water temperature, 

 as presented previously (Fig. 3B), exhibited a strong 

 inverse relationship with egg stage duration. We used 

 stepwise regression to build a model to predict egg-stage 

 duration as a function of egg size and temperature: 



^ U.S. Naval Observatory Astronomical Applications 

 Database. 2002. Table of sunrise/sunset. Astronomical 

 Applications Dept., Data Services. Website: http://aa.usno. 

 navy.mil/data (accessed August 2002). 



Y = 27.96 (±1.69 SE) -i- 



29.05 (±1.55) D- 



1.28 (±0.04) DT 



(r2=0.58, df=826, P<0.001), 



