FISHERY BULLETIN: VOL. 81. NO. 1 



60 



50-- 



E 

 E 



CD 



E 



CD 



o 



40-- 



30-- 



20 



400 



500 



700 



800 



900 



600 

 Standard Length (mm) 



Figure 5. — Orbit diameter plotted against standard length for male tilefish from the U.S. east coast and the Gulf of Mexico 



Table 6. — Percent male tilefish classified to sample locations 

 by discriminant function analysis. 



To sample locations 



CO 



6 cd 6 



X LL X w 



cu cdco <"— <o = 

 5<u5x5« = => 



Q.d)CU<0 



-. E o t- o * 



*- 03 



c 



o>_ 

 o 



O 5 u to 



o o y o 



X U- c ^ 



re 



•o CD 





CD 



Ol o 



CD 



CD o CD -g 

 o ° o. 





re ^5 _ro o 

 co--*;--;-';- 



From 

 sample locations 



5O5050 go 

 O O O co 





< w < 



5 5 



Gulf of Mexico- 

 Campeche Banks 



Gulf of Mexico- 

 off Texas 



Gulf of Mexico— 

 off west Florida 



South Atlantic Bight— 

 off South Carolina 



Mid-Atlantic Bight- 

 west of Hudson Canyon 



Mid-Atlantic Bight- 

 east of Hudson Canyon 



Mid-Atlantic Bight- 

 east of Block Canyon 



Mid-Atlantic Bight- 

 west of Veatch Canyon 



N 



43 14 43 _7 



17 83 _6 



18 12 70 V7_ 



9 75 8 8 12. 



9 55 18 9 9 22 



4 15 23 23 19 16 26 



2 5 7 5 54 27 41 



12 4 4 17 50 13 24 



Table 7.— Percent female tilefish classified to sample loca- 

 tions by discriminant function analysis. 



To sample locations 



From 

 sample locations 



Gulf of Mexico- 

 Campeche Banks 



Gulf of Mexico- 

 off Texas 



Gulf of Mexico- 

 off west Florida 



South Atlantic Bight- 

 off South Carolina 



Mid-Atlantic Bight- 

 west of Hudson Canyon 



Mid-Atlantic Bight- 

 east of Hudson Canyon 



Mid-Atlantic Bight- 

 east of Block Canyon 



Mid-Atlantic Bight- 

 west of Veatch Canyon 



Mexico populations. In the discriminant function 

 analysis South Carolina samples of both sexes 

 classified correctly a high percentage of the time 

 but misclassification occurred to both Mid-At- 

 lantic Bight and Gulf of Mexico samples. The 

 variability in the pattern of morphological char- 

 acters can be accounted for by clinal variation in 



these characters, or, less likely, by two distinct 

 groups that are only weakly differentiated. The 

 interpretation of the morphological data may 

 also be hampered by the small samples for more 

 southern populations and the great distances be- 

 tween them. 

 Other life history data for tilefish in the Mid- 



48 



