Crabtree et al Age and growth of Albula vulpes 



445 



ratio tests were used to compare parameter estimates 

 for males and females (Kimura, 1980; Cerrato, 1990). 

 We used the outlier detection methods described by 

 Sokal and Rohlf ( 1995) to test the significance of two 

 perceived outliers from the age-length and otolith 

 weight-age relations. Dixon's test statistic for outli- 

 ers was used for sample sizes <25, and Grubb's test 

 statistic for outliers was used for larger sample sizes. 

 Outliers were considered significant at the P<0.05 

 significance level. Because we could not sex YOY 

 bonefish, they were not included in growth models, 

 but they were included in length-length regressions 

 and pooled length-weight regressions. Sex-specific 

 length-weight regressions were calculated by linear 

 regression of log 10 -transformed data, and we com- 

 pared the slopes and elevations using a t-test (Zar, 

 1984). 



We used a catch curve to estimate the instanta- 

 neous total mortality rate (Z) from age-frequency 

 data (Robson and Chapman, 1961). The age of full 

 recruitment to the fishery for the catch-curve analy- 

 sis was assumed to be 5 years, and year classes rep- 

 resented by fewer than five individuals were elimi- 

 nated from the analysis. We estimated the instanta- 

 neous rate of natural mortality (M) using Pauly's 

 (1980) formula In M = -0.0152 - 0.02791n L m + 

 0.65431n K + 0.4631nT, where L m and K are param- 

 eters from the von Bertalanffy growth equation, and 

 T is the average annual surface temperature. An 

 average annual temperature of 25.9°C (SD=0.13) was 

 calculated from daily temperature measurements 

 made from 1992 to 1994 at the Keys Marine Lab- 

 oratory's water intake pump in Florida Bay. Pauly 

 (1983) suggested that for schooling species, such as 

 bonefish, the estimate of natural mortality derived 

 from the above equation should be reduced by 20%. We 

 used Pauly's formula and the 20% correction to sug- 

 gest a possible range of values for natural mortality. 



Undamaged otoliths were weighed to the nearest 

 0.01 mg in order to evaluate the use of otolith weight 

 as a predictor of age. If both left and right otolith 

 weights were available for an individual fish, the 

 mean of the two weights was calculated. Sex-specific 

 linear regressions were fitted to log 10 -transformed 

 otolith weight and age data, and we compared the 

 slopes and elevations of the regressions using a t- 

 testlZar, 1984). 



Results 



The 528 bonefish we examined ranged from 21 to 

 702 mm in length; 56 (10.6%) of these fish were YOY 

 (21-116 mm). Fish caught on hook-and-line gear 

 ranged in length from 228 to 702 mm (median=591 



mm, interquartile range=lll mm, n=452) and most 

 (80%) were from 500 to 700 mm (Fig. 2). Among all 

 bonefish that we sexed, females ranged from 228 to 

 702 mm in length (median=601 mm, interquartile 

 range=110 mm, «=274) and males ranged from 290 

 to 700 mm (median=566 mm, interquartile range= 

 146 mm, n=192). The size-frequency distributions of 

 males and females were significantly different 

 (Kolmogorov-Smirnov two-sample test, Z)=0.231, 

 P<0.001); females were significantly larger than 

 males (Mann-Whitney t/-test, P<0.001). Females 

 were also more abundant than males. In our sample 

 there were 192 males and 274 females, and the sex 

 ratio was significantly different from 1:1 <x 2 =14.43, 

 df=l, P<0.001). 



Neither slopes it-test, df=453, ^=0.474, P=0.318) 

 nor elevations U-test, df=454, ^=0.084, P=0.467) of 

 the length-weight regressions for male and female 

 bonefish were significantly different. The pooled 

 length-weight equation for sexed and unsexed fish 

 and the relations between SL, FL, and TL are pre- 

 sented in Table 2. 



Bonefish otoliths when viewed with reflected light 

 have clear and easily counted narrow, opaque (bright) 

 annuli that alternate with broad translucent (dark) 

 zones (Fig. 1). Proceeding from the otolith's core to- 

 wards the otolith's proximal margin, these translu- 

 cent zones become increasingly opaque in appear- 

 ance as the otolith grows. In the portion of the otolith 



