FISHERY BULLETIN: VOL. 81, NO. 1 



locus exhibited a monomeric pattern (heterozy- 

 gote was two banded) with fast and slow bands. 

 Our interpretation of the dimeric and monomeric 

 nature of these enzymes is consistent with past 

 studies of their molecular structure (Manwell 

 and Baker 1970). Distribution of EST and IDH 

 electromorphs was not significantly different 

 than expected from Hardy-Weinberg equilibri- 

 um (Tables 1 , 2), which is additional support for a 

 single-locus, two-allele genetic model. However, 

 it should be noted that the chi-square test is not 

 very sensitive at small sample sizes(<200)(Fair- 

 bairn and Roff 1980). 



There was no significant difference in geno- 

 type distribution as a function of length (age) for 

 both enzymes (EST X 2 = 1.16, P>0.6, n = 20; 

 IDH x 2 = 2.93, P>0.2, n = 20) in the sample ex- 

 amined (west side of Hudson Canyon). 



There were distinct patterns of variation 

 among the populations sampled (Fig. 3). Chi- 

 square contingency tests revealed no significant 



differences in genotype distribution within the 

 Mid- Atlantic Bight or southern sampling loca- 

 tions (South Carolina, west Florida, Texas, and 

 Campeche) (within Mid- Atlantic Bight EST X 2 5 , 3 

 = 8.77, 0.25<P<0.5; IDH x 2 5 , 3 - 9.45, 0.25<P< 

 0.5 : within southern locations EST x 2 4 , 3 = 4.12, 

 0.5<P<0.75; IDH x 2 3 , 3 = 7.34, 0.1<P<0.25). 

 However, differences in genotype distributions 

 between Mid- Atlantic Bight and southern sam- 

 ples were highly significant (EST x 2 9 , 3 = 45.01, 

 P<0.001; IDH x 2 8|3 = 111.76, P<0.001). 



Morphological Data 



Gill raker numbers were the only meristic 

 characters that were significantly different 

 among samples. All gill raker counts (upper, 

 lower, and total) for males and females from the 

 Mid-Atlantic Bight samples were not signifi- 

 cantly different from the west Hudson Canyon 

 sample (Tables 3, 4). However, total gill rakers, 



Table 1. — Comparison between observed genotypes and Hardy-Wein- 

 berg expectations (in parentheses) at the liver esterase locus for all 

 tilefish sampling locations. See Figure 3 for sample sizes. 



Niche = name applied by fishermen to area about 50 km east of Hudson Can- 



yon. 



ns = not significant. 



Table 2. — Comparison between observed genotypes and Hardy-Wein- 

 berg expectations (in parentheses) at the liver isocitrate dehydrogen- 

 ase locus for all tilefish sampling locations. See Figure 3 for sample 

 sizes. 



'Niche = name applied by fishermen to area about 50 km east of Hudson Can- 

 yon. 

 2 ns = not significant 



44 



