FISHERY BULLETIN: VOL. 84, NO. 4 



GULF OF MEXICO 



Figure 1.— Sampling areas for king mackerel, Scomberomorus cavalla, in the Gulf of Mexico and Atlantic Ocean during 1977-78. 



from a length-weight regression (r = 0.996; n = 

 186) of king mackerel data from all areas. 



Egg size distributions within the ovary were sta- 

 tistically compared to ensure that subsamples taken 

 for studies of maturation and fecundity were repre- 

 sentative (Yuen 1955; Otsu and Uchida 1959). Both 

 ovarian lobes were divided into three sections 

 (anterior, middle, and posterior) of about equal 

 length. At a selected point along each of these sec- 

 tions, a 2-4 mm thick cross section was cut and 

 removed. A wedged-shaped portion was then taken 

 from each of the three cross sections and divided 

 into three zones: inner, middle, and outer. A sam- 

 ple of 150 yolked eggs from each of the zones was 

 examined with a microscope and all eggs were mea- 

 sured to the nearest 0.02 mm at 500 x on whatever 

 axis the egg happened to be located in respect to 

 an ocular micrometer scale (Clark 1934). A chi- 

 square test of independence (Steel and Torrie 1960) 

 was used to test for significant differences in mean 

 egg diameters (EDs) among the sections, zones, and 

 zones within a section in each lobe. 



Each wedge-shaped sample of eggs was placed in 

 a dish with 10% Formalin 3 and the eggs were then 

 teased apart. Samples containing only unyolked 

 eggs (<0.20 mm ED) were considered to be from 



immature fish and only 100 eggs from these samples 

 were measured. Samples with yolked eggs (^0.20 

 mm ED) were considered to be from mature fish and 

 300 eggs were measured. 



Seasonal maturation was determined by plotting 

 monthly mean EDs of the most advanced eggs found 

 in each ovary and by gonadosomatic indices (GSI = 

 the percentage of TW represented by gonad 

 weight). The range and 95% confidence interval of 

 the monthly mean GSIs were also plotted. To com- 

 pare the variation of GSIs, we calculated the coef- 

 ficient of variation for each month. We estimated 

 the length at which the fish first matured by com- 

 puting mean GSIs for fish in each 50 mm interval 

 and used the length at which the greatest increase 

 in mean GSIs between consecutive FL intervals oc- 

 curred. For this analysis we only used data that were 

 collected during the fish's most sexually active 

 months as indicated by the highest values of mean 

 EDs and GSIs. An additional estimate was made for 

 females by assigning immature or mature status to 

 each fish according to egg stage and then calculating 

 the percentage of mature fish by FL intervals. 



3 Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



842 



