Ovaries from 186 females caught from 1972 

 through 1976 were examined. Fresh ovaries were 

 either blotted dry and weighed in grams or stored 

 in 107f Formalin'^ and weighed later. No sig- 

 nificant difference was found between the mean 

 weight of fresh and preserved ovaries (F = 0.0001; 

 df = 1, 16;P>0.75). The gonosomatic index (GSI), 

 ovary weight as a percentage of total body weight, 

 was used as an indicator of maturity. 



Preserved eggs 0.56 mm in diameter and larger 

 were counted when estimating fecundity. Eggs 

 <0.56 mm were less spherical in shape and in an 

 earlier stage of development. The 0.56-mm size 

 was determined by measuring the diameters of 

 3,912 eggs from mature, partly spent, and spent 

 fish. Small transparent ova were stained with 

 aceto-carmine to facilitate measuring. Egg 

 diameters were measured with an ocular microm- 

 eter at 30 X magnification and the orientation of 

 egg diameters was assumed to be random. Thin 

 cross sections were taken from the anterior, mid- 

 dle, and posterior parts of one ovary of a mature 

 fish and subdivided into three subsamples, repre- 

 senting the center, midregion, and periphery of 

 the ovary (Otsu and Uchida 1959). 



Fecundity was defined as the potential number 

 of mature eggs (yolked ova in the most advanced 

 size mode) that could be spawned during one re- 

 productive season and was estimated using a dry 

 weight method. For six fish in which entire ovaries 

 were saved for fecundity analysis, subsamples 

 consisted of a thin cross section taken from the 

 anterior, middle, and posterior parts of each ovary. 

 The eggs in these subsamples were separated from 

 the ovarian tissue, enumerated, dried, and 

 weighed according to the procedure described by 

 Baglin (see footnote 2). For six other fish, only the 

 ovary weight and a single cross section from the 

 middle of the ovary were taken; these cross sec- 

 tions comprised the subsamples. The eggs were 

 separated, counted, dried, and weighed. A dry/wet 

 weight regression was used to estimate the total 

 dry weight of the eggs in these ovaries, which were 

 not saved. Before the eggs in the subsample were 

 counted, 25 eggs 0.30 mm and larger from the two 

 most advanced modes were randomly selected and 

 measured. Eggs in this second most advanced 

 mode were included to give an indication of the 

 percentage of eggs in both modes because future 

 histological studies may indicate that these 



smaller eggs undergo further development and 

 are also spawned. Fecundity estimates, rounded to 

 the nearest 0.1 million eggs, were calculated from 

 the relationship: C = iAD/B) + A, where A is the 

 number of mature ova in the subsample, B is the 

 weight of the ova in the subsample, C is the 

 number of mature ova, and D is the weight of ova 

 from both ovaries. 



Results and Discussion 

 Sex Q)mposition 



From 1971 through 1977, sex was determined 

 from 1,128 white marlin from the gulf (Table 1). 

 The deviation from an expected 1:1 sex ratio was 

 significant from May through October. Sampling 

 was inadequate for the remaining months. 

 Females were more prevalent than males for each 

 month studied. 



From 1972 through 1977, sex was determined 

 for 720 white marlin from the Atlantic (Table 1). 

 There were 323 sex determinations from the 

 Florida Straits (March through May) and 397 from 

 the Middle Atlantic Bight (June through Sep- 

 tember). Sampling was inadequate from October 

 through February. No significant difference from 

 an expected 1:1 sex ratio was found for March, 

 May, July, August, and September, but a sig- 

 nificant difference was found for April and June. 

 For the months in which the sex ratio was sig- 

 nificantly different from the expected 1:1 ratio, 

 females were more prevalent. 



deSylva and Davis (1963) found a significant 

 difference from an expected 1:1 sex ratio (60% 

 females) when they combined their data from the 

 Middle Atlantic Bight for the summers of 1959 

 and 1960. They presented monthly sex composi- 



TABLE 1. — Monthly sex ratios for white marlin from the north- 

 ern Gulf of Mexico! 197 1-77), Florida Straits and Middle Atlantic 

 Bight (1972-77). 



^Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



'Significant departure from null fiypothesis at 0.05 level (chi-square) 



920 



