510 
Fishery Bulletin 11 5(4) 
75‘25'0*W 75*20'irW 75'15 - 0'W 75'10'0’W 75'5'CTW 
75 , '25'0* , W 75‘20’0'W 75'15 , 0’W 75 J 10‘0“W 75"5 , 0*W 
Maryland 
Maryland 
Atlantic Ocean 
Atlantic Ocean 
July CPUE 
• 0 04 0 15 
• 0 16 041 
• 042 • 0 56 
% 0 57 • 0 90 
A 091 - 1 34 
September CPUE 
• 0 00 - 0 08 
• 0 09 - 0 22 
• 0 23 - 0 49 
^ 0 50-071 
A 0 72-092 
75'25'0‘W 75*20*0*W 75'15'CTW 75'10'CTW 75“5‘CTW 
75'25'trw 75'200'W 75*15'0*W 75’10’0‘W 75'5’0‘W 
Figure 5 
Spatial distribution of black sea bass (Centropristis striata) in Maryland 
coastal bays in (A) May, (B) July, (C) September, and (D) October, derived 
from a time series of catch per unit of effort (CPUE) determined from trawl 
surveys conducted monthly during 1989-2013. 
Results from the GLMs show that average salin¬ 
ity at the sampling sites and the annual NAO index 
(AIC=90.07) are the most informative predictors of YOY 
abundance in the MCBs each year, and that salinity 
is the only significant predictor variable (Tables 1 and 
2); the annual NAO index alone was not an informa¬ 
tive predictor of age-0 catch (AIC=109.7). Catch of 
age-0 black sea bass was affected positively by salinity 
(P=0.0015) and negatively by the annual NAO index 
(P=0.0468). 
Model results showed no trend in the residuals (Fig. 
8) and a chi-square goodness-of-fit test showed no sig¬ 
nificant difference (P=0.166) between the observed and 
predicted values (Fig. 9). Owing to possible overdisper¬ 
sion, a negative binomial and zero-inflated model were 
run to compare AIC values from the Poisson model and 
the negative binomial and zero-inflated model. Com¬ 
parisons of AIC values indicated that the GLM with a 
Poisson distribution (AIC=90.07) had the lowest AIC 
value and fitted the data best when compared with the 
2 other models (negative binomial AIC=92.07, zero-in¬ 
flated AIC=92.26). 
Discussion 
There was no significant increasing or decreasing trend 
in abundance of juvenile black sea bass from 1989 to 
