388 



Fishery Bulletin 101(2) 



30 



25 



i. 20 

 2- 



c 



<n 'J 



E 

 o 



OQ 



5- 







E 

 o 



30-1 



25 



p 



■^ 20 



15 



10 



5- 



CAPE FEAR ESTUARY 



18 1 15 29 13 27 10 

 Mar Apr Apr Apr May May Jun 



30 



25 



20 



15- 



10 



5 







30 



25 



20 



15 



10 



5 



PAMLICO SOUND 



lower 



18 1 15 29 13 27 10 



Mar Apr Apr Apr May May Jun 



Figure 2 



Mean bottom water temperature and salinity for general areas (circle=lower, triangle= middle, x=mid-lower, 

 star=upper) of the Cape Fear and Pamlico Sound, NC, estuaries by sampling week, March-June 1987. Vertical bars 

 represent 95 percent confidence intervals around the means. 



areas and systems by using pairwise /-tests for all possible 

 pair combinations as an option in the general linear models 

 procedure (SAS Institute, 1988). 



Instantaneous and daily mortality rates were calculated 

 for spot and Atlantic croaker by using methods similar to 

 Crecco et al. (198.3) and Essig and Cole (1986). Inverse 

 regression (Zar, 1984) of the relationship between age 

 and SL was used to estimate age from SL. Ages were then 

 calculated for all fish sampled, and the natural logarithm 

 of the slope of the descending limb of the catch curve was 

 the instantaneous natural mortality rate (Z). Analysis of 

 covariance (covariate=age) in a general lineal model pro- 

 cedure was used to test for differences in mortality rates 

 between areas (SAS Institute, 1988). The daily mortality 

 rate (A/, '/(/d) for the whole time period was calculated as 

 Af =l-e2(Ricker, 1975). 



Because null hypotheses of no difference.s in growth or 

 mortality rates between areas were accepted in many cases, 

 probabilities of type-II errors (B, //„ actually false) existed. 

 Calculations of power (1-B), the probability of correctly 

 rejecting //„, are difficult with ANOVA or ANCOVA (Zar, 



1984; Neter et al., 1985). A power analysis was probably un- 

 necessary for spot and Atlantic croaker age-SL regressions 

 because the precisions of the slope estimates were good (i.e. 

 proportional SEs of the estimates were 2^% of the slopes). 



Results 



Hydrographic data 



Only bottom hydrographic data are presented because the 

 waters of these shallow stations were well mixed. Bottom 

 water temperatures in all areas of both systems were 

 similar on a given sampling date (Fig. 2). In the Cape Fear 

 estuary, mean water temperatures were not significantly 

 different «-test, P>0.05) between areas. Mean tempera- 

 tures throughout the Pamlico system were not significantly 

 different (P>0.05), except that the lower area was cooler 

 than the others (P<0.05). Comparisons between systems 

 revealed no significant differences (P>0.05) between middle 

 or upper area temperatures. Lower Cape Fear creeks were 



