Murphy and Taylor Scisenops ocellatus in Florida waters 



533 



Beamish 1980). A total of 100 oocytes in a randomly 

 chosen lamella were measured in each female gonad 

 to calculate mean oocyte diameters within subsamples. 

 To further delineate the spawning season and char- 

 acterize variation in egg size, ten of the largest oocytes 

 from each section were also measured (DeMartini and 

 Fountain 1981), and their mean was plotted against the 

 collection date. 



Otoliths were excised from subsampled fish, cleaned, 

 and stored dry in vials. Sagittae were sectioned in the 

 laboratory using a Beuhler Isomet Low-Speed Saw 

 with diamond wafering blades. A 0.5-mm-thick section, 

 cut through the core of the right sagitta, was mounted 

 on a microscope slide with Coverbond Mounting Media. 

 Each section was examined for age marks using a 

 dissecting scope (32 x ) with reflected light. All sections 

 were independently read once by two individuals; then, 

 when necessary, a common reading was conducted. 

 Determining age by simply counting annuli was not 

 always possible because annulus formation occurs 

 about 2-6 months after the anniversary of the actual 

 hatching date (see "Age Determination and Growth" 

 section). To assign ages accurately, we assumed a 

 biologically realistic hatching date of 1 October (see 

 "Spawning Size, Age, Season, and Location" section; 

 Matlock 1984), and ages were incremented on this date. 

 Observations of otolith sections from oxytetracycline 

 (OTC)-marked fish served to directly validate opaque 

 bands as annuli. Oxytetracycline-injected fish (intra- 

 muscular injection of 25 mg/kg body wt.) were held in 

 outdoor ponds (n = 4) for subsequent recapture and ex- 

 amination after they were free for more than 1 year. 

 All four fish were recaptured after 19 months. 



Red drum captured by all gear types were included 

 in the growth analyses. The von Bertalanffy (1957) 

 growth equation, FL = L^ (1 - exp [-K(t - /,,)]), was 

 fit to observed age-length data by using the NLIN pro- 

 cedure (Marquardt option) of SAS (Vaughan and Kan- 

 ciruk 1982). Growth equation parameters are defined 

 as follows: L^ = the average fork length (mm) that a 

 fish would achieve if it were allowed to grow indefinite- 

 ly in accordance with the model; K = Brody's growth 

 constant; t^ = the hypothetical age at which a fish 

 would have zero length, and t = age in years (Ricker 

 1975). The weight-length relationship was described by 

 linear regression of logm -transformed data. Analysis 

 of covariance was used to test for differences between 

 regressions. 



Age-length keys were applied to length frequencies 

 to separate age groups and develop sample abundance 

 data used to estimate total annual mortality rate (A): 



Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service. NOAA. 



A = 1 - S, where S is the annual survival rate. Age- 

 length keys were developed for each season for com- 

 bined samples captured by trammel net, hook-and-line, 

 and haul-seine. Abundance of each age-group in the 

 samples was then summed over all seasons during the 

 study period. Total annual mortality rates were esti- 

 mated from truncated age data that included all fully 

 recruited age-groups with sample abundance of five or 

 more fish. This procedure eliminated any bias inherent 

 in the collection of large fish (Chapman and Robson 

 1960). Survival rates were calculated using the modi- 

 fied Heincke method (Seber 1973), Robson and Chap- 

 man (1961) estimates developed for truncated age data, 

 and catch curves. 



Results and discussion 



Spawning size, age, season, and location 



Males matured at smaller sizes (x~-test, jO<0.05) and 

 younger ages than did females on both coasts. Some 

 Gulf coast males were sexually mature after they 

 reached 400 mm, and some on the Atlantic coast were 

 mature after they reached 350 mm (Table 2). Inter- 

 polated lengths at 50% maturity were 529 mm and 511 

 mm, respectively. Most males were mature at age 1 

 or 2, and all were mature by age 3 (see "Age Deter- 

 mination and Growth" section). Some Gulf coast 

 females were sexually mature after they reached 600 

 mm, and some on the Atlantic coast were mature after 

 they reached 550 mm. Lengths at 50% maturity were 

 825 and 900 mm, respectively. Some females were 

 mature at age 3, and all were mature by age 6. 



While results of previous investigations suggest that 

 the size and age at which red drum mature may vary 

 over its geographical range (Pearson 1929, Gunter 

 1950, Miles 1951, Overstreet 1983, Music and Pafford 

 1984), we show that there are also large maturational 

 differences between sexes. Lengths at 50% maturity 

 were about 300-390 mm smaller for males than 

 females. In Mississippi, males and females began 

 developing when they were 300-549 mm SL (358-629 

 mm FL); however, only after growing to 700 mm SL 

 (792 mm FL) were more than 50% of the samples of 

 either sex mature (Overstreet 1983). Female red dnim 

 in Texas matured during their fourth or fifth year, at 

 about 750-850 mm TL (71 1-803 mm FL) (Pearson 

 1929, Miles 1951, Matlock 1985). A 755-mm TL (716 

 mm FL) male red drum sampled in Georgia had well- 

 developed gonads and was apparently mature (Music 

 and Pafford 1984). 



Spawning peaked on both coasts of Florida from 

 about September through October. Maximum oocyte 

 diameters and the presence of eggs with yolk vesicles, 

 globules, and migrating nuclei indicated that active 



