FISHERY BULLETIN: VOL. 74. NO. 4 



for the fleet to become more dispersed as the 

 season progressed. 



Aggregation of the Boats 



The index of aggregation used in this study was 

 the mean separation distance of boats within a 

 specified area. The index was determined by sum- 

 ming separation distances between all boats in the 

 area and dividing this sum by the number of 

 separation distances. This calculation required 

 converting LORAN coordinates (given as the 2100 

 h PDT boat positions) to latitude-longitude coor- 

 dinates. Accuracy of the iterative technique used 

 to compute the coordinates has been estimated at 

 10 m (Thomas 1965:7-9, 38-52), although the absolute 

 position accuracy varied considerably due to the 

 precision of the LORAN operator and the distance 

 from the LORAN transmitters. Boat positions 

 reported at 2100 h within 200 miles of the coast are 

 estimated to be within 3 miles of the absolute 

 positions. 



Hunter (1966) stated that mean separation 

 distance is preferred for measuring relative 

 changes in spacing, but for comparison of samples 

 containing different numbers of individuals, mean 

 distance to nearest neighbor (Clark and Evans 

 1954) should be used. We did not use mean distance 

 to nearest neighbor because most fishermen fish 

 together with one or more companion boats. Mean 

 distance to nearest neighbor would thus represent 

 the average distance separating the same groups 

 of boats and would give little if any information on 

 actual compactness of the fleet within a specified 

 area. 



RESULTS AND DISCUSSION 



Fishing Power Versus Boat Length and 

 Number of Lines 



Sixty-six area-date strata (1° latitude by 1° 

 longitude rectangles and 1-day periods) were 

 selected to examine the relationship between the 

 fishing power of a boat and its length and number 

 of lines trolled. All strata had at least 20 boats 

 reporting within them. (The new mobile stratum 

 was not used here because the intent was to 

 partition the fishery area into a number of equal 

 quadrats, the size and location of the quadrat 

 being of no consequence. Daily boat positions had 

 been assigned to 1° longitude rectangles by 

 FPOW, so this stratum was used for convenience.) 



Fishing power estimates were then regressed on 

 boat length and number of lines. (Data on number 

 of lines were available only for the 1970 season.) In 

 none of the strata, in any season, was a significant 

 regression (F-test, P <0.05) found. This indicated 

 that no significant relationship existed between a 

 vessel's fishing power and its length or reported 

 number of lines trolled within a given 1° by 1° 

 rectangle during any given day. 



Because of the scatter of data for small-scale 

 time and area strata, the above conclusion did not 

 rule out the possibility of a significant relationship 

 between fishing power and boat length or number 

 of lines. Therefore, a larger stratum was chosen 

 which included all data for each year. Fishing 

 power estimates were again regressed on boat 

 length (1968, 1969, 1970) and number of lines 

 (1970). The results are shown in Table 2. 



Boat length was significantly related (P<0.05) 

 to fishing power of albacore boats in a time-area 

 stratum of one season and the entire fishery, 

 particularly in 1968. The significance of boat 



Table 2.-Regression equations and analysis of variance data for 

 boat length (in meters) and number of lines (1970) versus boat 

 fishing power. 



** significant at the 0.01 level. 

 • significant at the 0.05 level, 

 ns nonsignificant. 



976 



