Fischer et al.: Age, growth, mortality, and radiometric age validation of Lut/anus gnseus 



51/ 



ported maximum ages of 23 and 25 years for males 

 and females, respectively; the oldest fish in the study 

 was actually sampled from Grand Isle, LA. Burton 

 (2001) reported a non-sex-specific maximum age of 24 

 years. Sampling for both of these studies was focused in 

 Florida where there is higher fishing pressure on gray 

 snapper (Burton, 2001) and this fishing pressure may 

 explain the lesser maximum ages and paucity of older 

 individuals in their sample populations. 



Gray snapper exhibit multimodal distributions in 

 age and YOB frequencies. Due to minimum size limits, 

 very few individuals were represented below age 3. 

 Age distributions exhibited an initial peak at 3 years, 

 when gray snapper are beginning to recruit to the rec- 

 reational fishery. Successive peaks in age-class abun- 

 dance in our data set occurred every two years. In an 

 examination of abundance by YOB a similar pattern 

 was observed; strong year classes were followed by di- 

 minished year classes. Similar patterns of variability in 

 year-class strength have been observed in black drum 

 (Pogonias chromis) and red drum (Scienops ocellatus) 

 in the northern GOM. Beckman et al. (1989) suggested 

 that year-class variability in these species might be 

 due to environmental factors during early life stages 

 or biological controls on the population. If this observed 

 consistent pattern is reflective of the gray snapper popu- 

 lation off Louisiana, we suggest that the variation in 

 year-class strength may be reflective of intra-species- 

 specific year-class competition of juveniles competing 

 for resources within the estuaries before recruiting to 

 the offshore fishery. 



Researchers continually search for effective, cost-ef- 

 ficient ways to acquire fish age data. Body size has been 

 shown to be a poor value to use for estimating age in a 

 number offish species because of the considerable vari- 

 ability in size at age. Otolith growth has been shown 

 to continue with age, independent of somatic growth. 

 Otolith weight (W ) has been used as a predictive tool 

 to determine age in a number offish species (Temple- 

 man and Squires, 1956; Beamish, 1979; Wilson and 

 Dean, 1983; Secor et al., 1989; Beckman et al., 1991). 

 Although a strong relationship has been demonstrated 

 between W and age, especially for the younger age 

 classes, considerable variability exists in W at age in 

 older age classes. For example, the W of a 10-yr-old 

 male gray snapper can range from 180 mg to 357 mg 

 thus preventing a precise age estimate based on W 

 alone. Although W (l data may provide general informa- 

 tion on overall age distribution patterns of a popula- 

 tion, we feel that annulus counts from otolith cross 

 sections provide the most accurate age estimates for 

 gray snapper. 



Our overall (sexes combined) von Bertalanffy growth 

 model estimated a maximum theoretical length (Lj 

 of 656.4 mm TL. Although a likelihood ratio test indi- 

 cated a significant difference between male and female 

 models, this difference may be of limited biological 

 significance because male and female models appear 

 to be very similar. The presence of larger, older fish 

 in our sample population resulted in our overall model 



coming to an asymptote at a smaller L, and having a 

 larger respective k than previously reported (Manooch 

 and Matheson, 1981; Johnson el al . 1994) Johnson 

 et al. (1994) predicted an L r of 792.25 mm using the 

 regression method of Manooch and Matheson (1981) 

 to back calculate lengths at age. Johnson et al. (1994) 

 also obtained a much smaller estimate of ft at 0.08 

 compared with a k value of 0.22 predicted in our model. 

 A smaller estimate was not unexpected given the in- 

 verse correlation between L x and k noted by Knight 

 (1968). Because of the minimum size limitations on 

 the recreational fishery, smaller (presumably younger) 

 individuals below 304 mm TL were almost absent in 

 our sample population. We chose to not specify a y-in- 

 tercept for t and to force our growth models through 

 zero in order to obtain more accurate estimates of k. 

 Forcing our models through zero also contributed to 

 the differences in growth parameters between our study 

 and those of Johnson et al. (1994). Like Johnson et al. 

 (1994), Burton (2001) also estimated growth param- 

 eters by fitting back-calculated lengths at age. Burton's 

 (2001) L x estimates of 717 mm and 625 mm for north 

 and south Florida, respectively, are similar to those 

 found in our study. Burton's (2001) sample populations 

 consisted of a number of fish below 200 mm TL. These 

 smaller individuals had similar effects on his models 

 as that of forcing our models through zero. Burton's 

 estimates of k were 0.17 and 0.13 for north and south 

 Florida, respectively, compared with a /; of 0.22 for our 

 overall model. 



We estimated total instantaneous mortality (Z) to 

 be 0.17 and full recruitment to the fishery at age 4. 

 We chose to use the truncated age range of 5-16 years 

 (versus 5-28 years) for Z estimation in order to have at 

 least 10 samples in each age category. Our estimation 

 of Z based on all age categories (5-28) was 0.18. Our 

 estimate of Z is at the low end of the range of values 

 reported by Johnson et al. (1994) (Z=0.17-0.26) for the 

 Gulf of Mexico. It should be noted, however, that John- 

 son et al. (1994) pooled fish from five distinct geographi- 

 cal locations. Of the 432 fish analyzed in their study, 

 69% came from Grand Isle, LA (n = 104) and Panama 

 City, FL (n = 193). The remaining 31% came from the 

 central and southern coasts of Florida. Perhaps John- 

 son et al.'s (1994) estimates of Z would be lower if only 

 the Louisiana samples were used. Our Z values, how- 

 ever, are much lower than those reported by Manooch 

 and Matheson (1981) (Z = 0.39-0.60) and Burton (2001) 

 (Z=0.34-0.95) for the east coast of Florida. 



Our low estimate of Z for gray snapper in Louisiana 

 waters is clearly associated with the abundance of older, 

 larger individuals in the population. Unlike the catch 

 curves in previous studies that dealt with gray snapper 

 populations on the east coast of Florida (Manooch and 

 Matheson 1981; Burton 2001) and in the southeast in 

 general (Johnson et al. 1994), the mode of our catch 

 curve is not well defined. It is evident that gray snap- 

 per in the South Atlantic are heavily exploited (Burton, 

 2001), as evidenced from their age-frequency distribu- 

 tion and high estimates of Z. 



