Fischer et al.: Demographic structure of Lut/anus campechanus in the northern Gulf of Mexico 



597 



opaque annulus formation for this state. However, mini- 

 mum proportions of opaque edges during the months of 

 April through October may indicate that red snapper 

 from Texas form an opaque annulus during the winter 

 months. Proportions of opaque edges for Alabama and 

 Louisiana were essentially the same: maximum propor- 

 tions of opaque edges during the months of February 

 and March followed by a decrease to minimum propor- 

 tions during the months of May through November 

 (Fig. 3). These findings are consistent with previous 

 age and growth studies on red snapper in the northern 

 COM (Patterson et al, 2001; Wilson and Nieland, 2001), 

 indicating that the formation of one opaque annulus in 

 the winter months is followed by the formation of one 

 translucent annulus in summer. Annulus-based age 

 estimates of red snapper from the northern GOM have 

 also been validated to 55 years with otolith radiocarbon 

 chronologies based on accelerator mass spectrometry 

 14 C measurements (Baker and Wilson, 2001). 



Red snapper ages ranged from 1 to 45 years and the 

 majority (90%) of individuals were between 2 and 6 

 years (Fig. 4). Alabama fish ranged from 1 to 35 years 

 (re=1985), Louisiana fish ranged from 2 to 37 years 

 (n=1864). and Texas fish ranged from 1 to 45 years 

 (rc=1186). Modal ages were 4 years for Alabama and 3 

 years for Louisiana and Texas red snapper. We found 

 significant differences among age-frequency distribu- 

 tions from all states (AL and LA, MD = 9; AL and TX, 

 MD = 33.84; and LA and TX, MD=24.84). Texas had a 

 much higher proportion of younger individuals; 63% of 

 sampled fish were aged at 3 years or less compared to 

 only 30% of Alabama and 39% of Louisiana fish aged 

 at 3 years or less. 



Red snapper growth was modeled from weighted mean 

 FL at age and mean TW at age by using the von Berta- 

 lanffy growth equation (Fig. 5, A and B). Resultant von 

 Bertalanffy growth equations were 



ALFL. = 839(1 -e 1 - 0381 " 1 ) 



(F, 15 = 2824.9; P<0.0001; r 2 =0.95); 



LAFL„ = 847.8(1 - e'-° • 25 "") 



(F 1; 13=5024.4; P<0.0001; r 2 =0.76); 



TXFL„ = 778.2(1 - e <-o.49tt)>) 



(F 1;19 =1452.1; P<0.001; r 2 =0.85); 



AL TW 



17.05(1 -e 



(-0.15inii3.03 



(F 1;15 =457.9; P<0.0001; r 2 =0.89); 



LA TW_._ = 12.61(1 - e c-o.32(»))3.03 



(F 114 =122.02; P<0.0001; ;- 2 = 0.18); 



TX TW rr , = 8.89(1 - e'- - 21 "") 2 84 



(F 1;12 = 613.01; P<0.0001; r 2 =0.96). 



Models of mean red snapper FL at age for Alabama and 

 Louisiana were markedly similar with likelihood ratio 

 tests indicating no significant differences between red 

 snapper from the two states (Table 3). However, the 



Texas model differed from both Alabama and Louisiana 

 models. The Texas model displayed significant differ- 

 ences from the other models in both L m and in k. A 

 comparison of the models of mean TW at age indicated 

 no significant differences between Alabama and Loui- 

 siana red snapper (Table 3). Differential growth in TW 

 was found when comparing Alabama and Louisiana with 

 the Texas model; significant differences were manifested 

 in both W M and in k. The model failed to converge for 

 estimating a common value of k for both Louisiana and 

 Texas. 



We recognized that the larger red snappers from 

 Louisiana might bias the data; therefore we compared 

 growth for fish from 2 to 10 years of age — a time pe- 

 riod when red snapper have demonstrated rapid linear 

 growth (Szedlmayer and Shipp, 1994; Patterson et al., 

 2001; Wilson and Nieland, 2001). Linear regressions 

 of mean FL at age for all individuals 2 to 10 years 

 (Fig. 6A) were compared among states. We found no 

 significant differences among states (ANCOVA test of 

 homogeneity of slopes, F 2;28 =2.7; P=0.08; ANCOVA test 

 for equal intercepts, F 2 . 28 =0.52; P=0.6). 



