McBride et al.: Spawning grounds of Hemiramphus brasiliensis and H. balao 



585 



essary. Weights offish collected from July 1997 to October 

 1998 were measured for up to 30 females per species per 

 trip. Fish body and gonad weights were only occasionally 

 recorded for other trips during 1995-99, but these data 

 were included in the mapping of ripe females (i.e. females 

 with hydrated oocytes in ovigerous lamellae) to increase 

 overall sample size. In total, weight data were collected 

 for 2908 halfbeak females from 79 commercial fishing net 

 sets (63 different fishing days) and 59 fishery-independent 

 sampling events (50 different sampling days). Commercial 

 catches contributed 1649 ballyhoo and 757 balao females, 

 and fishery-independent collections added another 497 bal- 

 lyhoo and 5 balao females. The gonadosomatic index (GSI) 

 was calculated as 



GSI=(GWnBW-GW)) X 100, 



locations were recorded by using a global positioning sys- 

 tem hand-held unit. The latitude and longitude of fishery- 

 dependent samples were taken onboard the fishing vessel 

 once the lampara net enclosed the fish. Location data for 

 fishery-independent samples were taken from an anchored 

 position. Depth information was divided into the following 

 categories: area exposed at low tide, 0-1 m, 1-2 m, 2-4 m, 

 4-6 m, 6-10 m, and 10-20 m; only one of 159 locations was 

 without depth information. Substrate information was di- 

 vided into one of the following categories; platform margin 

 reefs, patch reefs, other hard bottom, seagrass beds, and 

 bare substrate; 32 of 159 locations did not have substrate 

 information. 



Results 



where GW = gonad weight; and 

 BW = body weight. 



The processes of final oocjrte maturation (FOM) were 

 examined by comparing GSIs with changes in whole 

 oocyte size and histological criteria. Oocyte diameters 

 were measured for 39 ballyhoo collected in April 1998 and 

 March-April 1999. Fixed ovary tissue was washed, teased 

 apart, and placed in a solution of 33% glycerin to 67% water 

 Measurements of at least 300 oocytes per fish were made 

 to the nearest micron with the aid of a video system and 

 image-analysis software (Optimas, vers. 100, Media Cyber- 

 netics, Inc., Silver Spring, MD). A minimum-size cut-off of 

 0.15 mm was used to exclude debris within the petri dish. 

 Initially, oocyte diameters from six ballyhoo were measured 

 from four separate sections of ovaries (left, right, anterior, 

 posterior), but the modal oocyte diameter within each in- 

 dividual was the same for all four sections; therefore tissue 

 from other fish was extracted without regard to location 

 within the ovary. Berkeley and Houde (1978) performed a 

 similar test and came to a similar conclusion. Ovaries from 

 fish {n=930 females) collected during March and May 1997, 

 July 1997-October 1998, and March-April 1999 were pre- 

 pared for histological examination. Histological methods 

 are presented in McBride and Thurman (2003). Here, the 

 most advanced oocyte stage was recorded (in increasing 

 order of oocyte maturity) as either perinucleolar, cortical al- 

 veolar, vitellogenic. or as either of two stages for oocytes in 

 final maturation; nucleus migration and hydration (West, 

 1990 ). McBride and Thurman ( 2003 ) have reported the size 

 at 50% maturity to be >160 mm FL for female balao and 

 >198 mm FL for female ballyhoo (approximately 31.5 gand 

 60.9 g, respectively, using length-weight relationships from 

 Berkeley and Houde [1978; their Fig. 71). Mean GSIs and 

 95% confidence limits were determined for fish with regard 

 to their most advanced stage of oocyte development, and 

 a minimum cut-off value was established for GSI values 

 indicating ripe females. 



The locations of ripe females were plotted to indicate 

 spawning grounds. Water depth and bottom type of these 

 spawning locations were determined by using the Marine 

 Resources Geographic Information System at the Florida 

 Marine Research Institute (w-ww.floridamarine.org). Point 



Spawning cycles 



Ballyhoo and balao had prolonged spawning seasons that 

 peaked in late spring and early summer (Fig. 2). Monthly 

 average GSIs of mature females increased from a low of 

 <0.4 for both species to a high of 6.4 for balao and 6.9 for 

 ballyhoo. The average GSI of individual females with only 

 primary growth oocytes (i.e. their most advanced oocyte 

 stages were perinucleolar or cortical alveolar) fell within a 

 narrow interval of 0.1-0.3 (maximum=0.95, Fig. 3). These 

 females were either small fishes that were immature or 

 they were larger fishes that were regressed (i.e. mature 

 but inactive). Vitellogenesis more than doubled the aver- 

 age GSI values for both species, but all females whose 

 most advanced oocyte stage was vitellogenic had GSIs less 

 than 1.37. Dramatic increases in GSI values also occurred 

 during FOM, and significant differences were evident in 

 the sequential FOM steps of nucleus migration and nucleus 

 breakdown. During nucleus migration, but before hydra- 

 tion, ballyhoo GSIs averaged 3.4 (3.3-3.6; 95% CD and 

 balao GSIs averaged 5.2 (4.6-5.9). Females with hydrated 

 oocytes had GSIs averaging 7.4 (7.0-7.9; 95% CD for 

 ballyhoo and 8.7 (6.6-10.8) for balao. Individual female 

 GSIs reached an observed maximum of 13.3 for ballyhoo 

 and 14.2 for balao. By applying these GSI criteria, which 

 indicate that females with a GSI greater than about three 

 had oocytes in FOM, it is evident that the average mature 

 halfbeak female is actively spawning from at least March 

 to August. 



Final oocyte maturation also followed a diel cycle. For 

 ballyhoo, FOM began about 30-36 hours before ovulation, 

 hydration of oocytes began about 8-12 hours before ovula- 

 tion, and ovulation occurred at sunset. Ballyhoo oocytes de- 

 veloped in a group-synchronous pattern, and during FOM, 

 a batch of oocytes increased rapidly in diameter (Fig. 4). 

 Mature female ballyhoo had a bimodal or trimodal distri- 

 bution of oocyte diameters when spawning. The smallest 

 mode (<1.0 mm oocyte diameter) represented a reservoir 

 of primary growth oocytes and vitellogenic oocytes. Larger 

 modes, between 1 and 3 mm, represented oocytes in FOM. 

 The presence of two larger ooctye modes (-1.0-2.0 and >2.0 

 mm) in all females sampled during the afternoon period in- 

 dicated that female ballyhoo typically spawn every day dur- 



