Tl'RNER ET AL: TILEFISH IN MIDDLE ATLANTIC-SOUTHERN NEW ENGLAND 



Hudson Canyon 



40 

 30 

 20 

 10 



1974 



Southern 

 New England 



n = 175 





■f* 



T I | (— ■— 1 3 



40 

 30 

 20 

 10 



1980 



n = 601 



~~i — « — i — i — I — ^—i — ' 

 40 60 80 100 



Length (cm) 



n -1587 



— pu, j , , ;-. r , 1 



40 60 80 100 110 

 Length (cm) 



FicURE 5. — Length frequency of tilefish caught by longline 

 in the Hudson Canyon and Southern New England areas, 

 1974-80. If there were 200 or more observations in a year 

 and area, the data were weighted equally by trips within a 

 season and by percentage of annual landings made in each 

 season; no weighting was used if there were <200 obser- 

 vations in a year and area, n = sample size. 



can's multiple range test: P = 0.05), and these dif- 

 ferences exceeded 1 cm FL after the fifth year (Fig. 

 6). Fish of unknown sex grew more slowly than 

 females through their fifth year (significantly slower 

 at ages 3-4, Duncan's test: P= 0.05), and they always 

 grew more slowly than males ( significantly so for ages 

 3-5, Duncan's test: P= 0.05). 



ANOVA and Duncan's multiple range test on last 

 back- calculated lengths for fish caught in different 



locations and by different fishing gears showed no 

 significant differences between areas (P> 0.05) and 

 indications of gear selection for only some ages. Sex 

 was included as an effect in the ANOVA to account 

 for differential growth rates. Foreign trawl and long- 

 line caught fish were equal or nearly equal (within 1 

 cm) in size at ages 3, 4, 6, and 7, when their average 

 lengths were 29, 41, 53, and 63 cm FL, respectively. 

 Recreationally caught fish were significantly larger 

 at age 4 (Duncan s test: P= 0.05) when they averaged 

 46 cm FL and at age 5 when all gears were significant- 

 ly different .(ANOVA: P < 0.01 and Duncan's test: P 

 = 0.05) with the recreational catch averaging 51 cm 

 FL, the longline catch being intermediate in size (49 

 cm FL), and the foreign trawl- caught fish being 

 smallest (46 cm FL). The recreational catch at age 7 

 was significantly smaller (Duncan's test: P = 0.05) 

 than the catch by the other gears (63 cm FL). In all 

 other years, differences in size were not significant. 



Growth Models 



Growth models (von Bertalanffy 1938) for males 

 and females were found to be significantly different 

 with Hoetelling's T 2 (P< 0.01). Curves were fit using 

 last back-calculated length at age, except for the 13 

 fish with aberrant otolith growth patterns for which 

 empirical length was used. Females had a much 

 smaller L x (90 cm FL) and a larger K (0.153) than 

 males (L„ = 111 cm FL and A' = 0.130). To describe 

 growth of the entire population, a von Bertalanffy 

 curve was also fit for all tilefish (Fig. 7). 



Ricker's (1975) population growth statistic (G) 



E 

 o 



c 



E 



<x> 



k_ 



o 



c 



12 



10 



8 



6 



4 

 2 



B 



a ® 





Male 



Female 



Unknown 



a 



 • 



• : 



a 



• . 



• • • 



a s 



a a a a 

 ® 



- - 



 • • 



10 15 20 25 

 Age (years) 



30 



FIGURE 6.— Annual growth increment of tilefish by age and sex. Data 

 for the first year may be inaccurate, because the back-calculation 

 formula was fit to tilefish that were at least 1 yrold. The mean of each 

 increment is shown by the point ('). 



757 



