FISHERY BULLETIN: VOL. 74, NO. 4 



were punched on computer cards. Each card con- 

 tained three pieces of information: the boat 

 number, an area-data code (signifying the 1° 

 latitude by 1° longitude rectangle and the calen- 

 dar day), and the boat's catch of the day. There 

 were approximately 3,300 observations in 1968, 

 1,500 in 1969, and 1,000 in 1970. 



A particular boat was chosen to represent the 

 standard unit of effort. Criteria for the standard 

 boat choice included the following: it fished 1) 

 during all three seasons; 2) in area-date strata 

 concurrently with a majority of the fleet; 3) most 

 of each season; and 4) consistently to provide a 

 standard, nonvarying reference for the other 

 boats. 



Estimates of fishing power"* of all boats in the 

 fleet were initially determined relative to the 

 standard boat. This was accomplished using a 

 computer program called FPOW (Berude and 

 Abramson 1972). FPOW utilizes Robson's (1966) 

 linear two-factor analysis model for estimating 

 the relative fishing power of fishing vessels. The 

 estimates of fishing power derived from the model 

 are logrithms. FPOW provides an approximate 

 correction for this bias using a Taylor series 

 expansion of the estimate about its true value. The 

 method and assumptions used in FPOW are de- 

 scribed in Robson (1966) and Abramson and Tom- 

 linson (1972:1022-1023). The program's storage 

 capacity was limited to 2,000 catch observations 

 from a combined total of not more than 200 

 distinct boats and area-date strata. Data for each 

 year were broken up into time segments short 

 enough to satisfy this limitation. Ten segments 

 were required in 1968, five in 1969, and three in 

 1970. Each segment was run independently and 

 provided estimates of each boat's relative fishing 

 power during the time segment. 



Considerable within-season variation occurred 

 in the average fishing power of the fleet (Table 1), 

 suggesting that the standard boat fished inconsis- 

 tently relative to the fleet. An examination of the 

 logbooks showed that the standard boat occasion- 

 ally experienced periods of very low catches (10 to 



^Fishing power is defined (Beverton and Holt 1957:172) as the 

 ratio of the catch per unit of fishing time of a particular vessel to 

 that of another vessel designated as the standard. It is assumed 

 that both boats must have fished on the same density of fish 

 during the same time interval and within the same fishing area 

 when the ratio is determined. Fishing success, on the other hand, 

 is related to fishing power but is more descriptive. It includes 

 parameters difficult to quantify. For example, fishing success 

 may include crew motivation, attitude, and access to useful 

 information. Together with fishing power, these parameters are 

 determinants of fishing success. 



15 fish per day) while the majority of the fleet in 

 the immediate area was catching 100 to 200 fish 

 per boat. This was particularly obvious during 

 segment 1 of the 1969 season. 



As a result of the standard boat's inconsistent 

 fishing, values of standardized catch per boat day 

 were also inconsistent between data segments. 

 For example, an average boat had fishing powers 

 of 3.70 and 1.01 on 25 July and 26 July 1969, 

 respectively (Table 1). If the average boat caught 

 100 fish on 25 July and 100 on 26 July 1969, values of 

 standardized catch per boat day (100 fish/average 

 fishing power) would be 27 and 99, respectively, for 

 these 2 days. Therefore a serial examination of 

 apparent abundance could not be performed 

 without normalizing fishing power estimates of 

 each boat in each data segment. 



Fishing power estimates were normalized by 

 subtracting the appropriate segment's average 

 fishing power from each boat's fishing power and 

 adding unity. (By definition the standard unit of 

 eff'ort is 1.0.) Each boat's fishing power estimate 

 was now relative to the average fishing power of 

 all boats fishing during the data segment. This 

 procedure required the assumption that the fleet 

 fished consistently relative to the standard boat 

 throughout each season. 



Daily standardized catch per boat within each 

 area-date stratum was determined by summing 

 the fish catches and dividing by the summation of 

 fishing power in that area-data stratum. The 

 standardized catch per boat day is an index of 



Table l.-Data segments for the 1968, 1969, and 1970 albacore 



seasons. 



974 



