FISHERY HI'LLKTIN: VOL. 8fi. NO. 4 



test indicated that maturity-at-age in males for the 

 two years was highly significantly different (Table 



4). 



Analysis of visual maturity at age data gave less 

 ambiguous results. Percent of females mature at age 

 in 1978 and 1982 was not significantly different 

 (Table 5). However, males matured at younger ages 

 in 1982 than 1978, and the differences between 

 years were highly statistically significant (Table 

 6). 



Table 5.— Female sexual maturity at age (visually determined) and 

 log-likelihood contingency tests that maturity at age in female 

 tilefish was not different for 1978 and 1982. Critical chi-square 

 values are x^o.os (9 df) = 16.9 and x^o oi (9 df) = 21 .7. NT = not 

 testable. 



1978 



1982 



Sex Ratio 



To minimize the chances of obtaining biased 

 results caused by the effects of fishing we estimated 

 sex ratio at size and age using only data collected 

 in 1978 when the tilefish population was lightly ex- 

 ploited. Proportions of males and females were 

 similar in the 46-50 and 51-55 cm FL intervals, al- 

 though the ratio in the 51-55 cm FL interval tested 

 significantly different at P < 0.05 (Table 7). Males 

 were significantly predominant (70-80%) between 

 71 and 90 cm FL. At the 91-95 cm FL size females 

 were predominant once again (>50%), although not 

 significantly so (Table 7). Above 100 cm FL only 

 males were collected. 



The general pattern of sex ratio at age in 1978 

 seems to have been equal proportions of the sexes 

 through about age 20 years, with only females pres- 

 ent from ages 29 to 36 years. Log-likelihood tests 

 of sex ratio at age showed no significant differences 



Table 6.— Male sexual maturity at age (visually determined) and 

 log-likelihood contingency tests that maturity at age in male tilefish 

 was not different for 1978 and 1982. Critical chi-square values 

 are x^o.os (^ df) = 15.5 and x^o.oi (^ ^^ = 20.1 . NT = not testable. 



1978 



1982 



Age 



(yr) 



Maturity 

 No. % 



Sample 

 size 



Maturity 

 No. % 



Sample 

 size 



G- 



statistic 



Table 7.— Sex ratio at length and log-likelihood tests that 

 sex ratio was not different from 1:1 at 5 cm FL intervals. 

 All G scores were calculated using Yates correction for 

 small sample sizes. Critical chi-square values are x^oos (^ 

 df) = 3.84 and x^ooi (^ df) = 6.64. NT = not testable. 



from equality except for ages 7 and 8 years; sex ratio 

 at age was not testable for ages 29-36 years because 

 only females were present (Table 8). 



Although sex ratio was skewed with age and size, 

 the estimated sex ratio for the entire population was 

 different from 1:1 in 1978 (Table 9). We calculated 

 the population to be 46.2% males, and could not re- 

 ject the null hypothesis that sex ratio is not differ- 

 ent from 1:1 (x^ = 0.15, df = 1, x"o.o5 = 3.84) in the 

 population >50 cm FL. 



Spawning 



Several lines of evidence suggest that tilefish in 

 the Mid- Atlantic-southern New England area are 



752 



