POLACHECK: DISTRIBUTION OF SEARCHING EFFORT 



with the largest distances between sets also have 

 large average search distances, but the converse 

 is not true (Fig. 5). The importance of dolphin 

 fishing as compared to log and school fishing ap- 

 pears not to be related to these differences among 

 cruises (Fig 6). It should be noted that the points 

 in Figure 5 must lie above the straight line with 

 a slope of 1.0. The expected relationship between 

 the variables in Figure 5 depends upon both the 

 underlying searching process and the spatial dis- 

 tribution of potential sets. While a positive rela- 

 tionship would be expected if both of these are 

 random, a more precise definition is beyond the 

 scope of this paper. The purpose in presenting 

 Figures 5 and 6 is to display the range of differ- 

 ences in the success of vessels in locating poten- 

 tial fishing targets. The large variation among 

 cruises suggests that all vessels may not be using 

 the same searching strategy and is an important 

 factor to keep in mind when considering the re- 

 sults from the clustering algorithm. 



^ 220 



1 



■5 200 



S 180 



u 



oc 

 o 



»- 



Ui 

 (A 



X 



u 



UI 



160 



140 



120 



O 100 



I- 



D 

 UI 



< 



UI 



<A 



80 



60 



< 



« 40 



UI 20„ 

 S 





_L 



_L 



10 



20 



30 



40 



_L 



50 



JL 



60 



70 



MEAN DISTANCE TO THE NEAREST 

 SET OR CHASE (nautical miles) 



Description of Clusters 



The clustering algorithm grouped between 30 

 and 100% of all sets and chases within a cruise 

 into clusters with at least three members (Table 

 1). The percentage of all sets and chases that are 

 included in clusters increases with the value of 

 the cluster parameter so that 60-100% of all sets 

 and chases within a cruise occur within 150 miles 

 of another one. These percentages can only in- 

 crease with increases in the value of the cluster 

 parameter (i.e., a set that is included within a 

 cluster for a lower value of the cluster parameter 

 will always be included in a cluster at a higher 

 value). Similarly, the average cluster size and the 

 total percentage of the distance searched that oc- 

 curs within clusters must be nondecreasing func- 

 tions of the clustering parameter (Table 1). How- 

 ever, even when the clustering parameter equals 

 150 miles so that most sets occur within clusters, 

 the percentage of the total distance searched 

 within clusters averages only 59% of the total 

 distance searched during the cruise. This indi- 

 cates that substantial searching activity occurs 

 far from any set or chase. 



The average intercluster distance for a cruise 

 ranged from 134 to 425 miles for clusters defined 



220 

 200 



z 



UJ -^ 



an 



< w 



Ui UJ 



«« 



liJ ^ 



O X 



ro 



^i 

 <S< 



o« 



Z H 



< UJ 

 UJ V> 



s 



180 



160 



140 



120 



100 



80 



60 



40 



20 

 



• . 



• • 



•• • 



'. 



10 20 30 40 50 60 70 80 



PERCENT OF THE SETS AND 

 CHASES INVOLVING DOLPHINS 



90 



Figure 5. — The relationship between the average distance 

 searched between sets and chases within a cruise and the aver- 

 age distance to the nearest one. 



Figure 6. — The relationship between the average distance 

 searched between sets and chases and the proportion of sets and 

 chases in which schools of dolphins were associated for a cruise. 



357 



