FISHERY BULLETIN: VOL. 85, NO. 3 



have biased CV (D) if the distribution of sightings 

 for the days were biased due to the effects of sea- 

 son or area. If we had included more large sam- 

 ples in our simulation, the linear relationship 

 may not have been evident. 



Calculation of K Values 



Because all terms used to calculate Var {N) 

 equal 0(l/n) and Var {N) is a linear sum of the 

 terms, Var (N) = Oil/n) or CV (N) = Oil/Vn). 

 Therefore, the relationship 



CY(N) ^KlVii 



(8) 



can be used to determine the change in CV (N) for 

 various values of n, where /C is a constant. This 

 relationship is true if the number of schools 

 sighted is proportional to population size. This 

 seems to be a reasonable assumption, although a 

 more complicated relationship between density 

 and school size, based on dolphin social structure 

 and its interaction with the fishery process, is 

 possible. K values for spotted dolphins in the 

 north and total areas were calculated for methods 

 A and B using the 1977-83 data (Table 1). These 

 K values were then used to determine CV iN )s for 

 specified values of n which would be expected 

 assuming from one to three annual ship surveys. 



Detection Rates 



The number of expected sightings with use of 

 one to three ships was calculated by computing 

 detection rates as the average number of dolphin 

 sightings per searching day. A day's searching 



effort generally consisted of searching from 

 sunrise to sundown; therefore, we assumed most 

 survey days covered approximately the same 

 trackline distance. However, distance searched 

 may vary inversely with rates of detecting dol- 

 phin schools because effort is halted so that ob- 

 servers can identify schools and make school size 

 estimates. The number of survey days, and hence 

 number of ships, required to obtain a specified CV 

 {N) was determined by dividing the number of 

 required sightings by the rate of detecting 

 schools. 



Detection rates were caculated separately for 

 data from the Jordan cruise and from the 

 Cromwell cruise because of the wide disparity in 

 detection rates of dolphins from the two vessels 

 when operating simultaneously in the calibration 

 area (Table 2). The Jordan has a much better 

 platform from which to detect dolphins because 

 its observation station was higher relative to the 

 water and because the Jordan rode much 

 smoother than the Cromwell. Pooled Jordan and 

 Cromwell detection rates were calculated by stan- 

 dardizing the Cromwell rates to Jordan rates 

 (Table 2) as 



DR = 



RjTj + R(.T(.C 



(9) 



where DR = pooled standardized detection rate 



for all dolphin schools, 

 Rj = dolphin schools detected per day by 



observers aboard the Jordan , 

 Re = dolphin schools detected per day by 



observers aboard the Cromwell, 

 T. = days searched aboard the Jordan , 



Table 2. — Detection rates of all dolphin schools from the Jordan and Cromwell in the calibration area and 

 pooled standardized detection rates for both vessels combined calculated in each stratum. Standardized 

 detection rates were calculated using the ratio of Jordan to Cromwell detection rates in the calibration area. 



442 



