FISHERY BULLETIN: VOL. 85, NO. 3 



CANAVERAL SHIP 

 CHANNEL SURVEY AREA 



N 



1 mile 



#SPOIL SITE 



Figure 2. — Description of the Port Canaveral ship channel survey area. The channel was separated into 1,483 m stations (7-14), 



which were divided into 30 m wide substations (A, B, C, and D). 



Var (Nq) = {SE)2 [l/« + X2/S(X. - X)^] 



where SE = standard error of the estimate pro- 

 vided by the straight line fit 

 n - sample size of the catch data set 

 X, = observed catch per sample in the i^^ 

 sample and 



The estimated variance of P was then calcu- 

 lated using procedures of Mood et al. (1974): 



Var iP) - (Ci/iVo^)2 Var iVo- 



In one instance, the experiment was conducted 

 450 



in an area larger than the standard substation 

 and a ratio (standard area/larger area = 0.75) 

 was used as a constant multiplier to standardize 

 estimates. 



The mean probability of capture was calculated 

 by combining all experimental P's using the for- 

 mulae: 



P = P,/k, and Var iP) = Var iP.Vk'^ 



where k = number of estimates. 



Once the efficiency of trawling equipment had 

 been determined, the number of loggerhead tur- 

 tles present in a substation (N) was estimated 

 using the following formula (Seber 1973): 



