ANALYSES OF THE RELATIONSHIP BETWEEN 



THE DISTRIBUTION OF SEARCHING EFFORT, TUNA CATCHES, AND 



DOLPHIN SIGHTINGS WITHIN INDIVIDUAL PURSE SEINE CRUISES 



Tom Polacheck^ 



ABSTRACT 



The fine scale distribution of searching effort within individual purse seine cruises in the eastern 

 tropical Pacific is analyzed in relationship to sightings of spotted dolphin, Stenella attenuata, and 

 tuna catches. The data for these analyses were derived from detailed observations made by National 

 Marine Fisheries Service observers aboard U.S. purse seiners. A clustering algorithm is developed 

 which separates the activity of a vessel into areas where sets were common and areas where they are 

 infrequent. Within clusters of high set densities, vessels tend to concentrate their searching effort. 

 Vessels searched proportionately greater distances relative to the physical distances between sets 

 while within clusters than when outside clusters. Encounter rates with schools of spotted dolphins 

 tend to be either much higher or much lower within defined clusters than outside them. Clusters with 

 low encounter rates were clusters in which non-dolphin associated tuna sets predominated. Because 

 of this dichotomy in the magnitude of the dolphin encounter rates within clusters, overall encounter 

 rates appeared to have relatively small biases if the concentration of searching effort within clusters 

 is ignored. The average catch of tuna per set was higher within the defined clusters than between 

 them. The overall results suggest that fine scale geographic effects need to be considered when using 

 data from purse seiners to examine changes in relative abundances of either dolphins or tuna. 



Catch and effort data underlie most indices of 

 abundances used for assessing the status of com- 

 mercially exploited fish stocks. The validity of 

 using catch and effort data from commercial har- 

 vests has long been questioned because of the 

 likelihood that fishermen concentrate their effort 

 in areas of high fish densities (Helland-Hansen 

 1909). Yet, almost nothing is known about the 

 allocation of effort by individual vessels. If data 

 are available on the activity of individual vessels, 

 these data have been mainly used for standard- 

 ization of effort. The catch and effort activities 

 within individual cruises have not been examined 

 in detail (in part because the data for such an 

 examination generally do not exist). The purpose 

 of the present paper is to examine the searching 

 behavior of tuna purse seiners in the eastern trop- 

 ical Pacific (ETP) based on detailed data compiled 

 by the National Marine Fisheries Service 

 (NMFS). The main questions addressed are 

 whether seiners concentrate their effort and what 

 is the relation between searching behavior, en- 



iSouthwest Fisheries Center La Jolla Laboratory, National 

 Marine Fisheries Service, NOAA, P.O. Box 271, La Jolla, CA 

 92038 and Department of Biology, University of Oregon, Eu- 

 gene, OR 97403; present address: Northeast Fishery Center 

 Woods Hole Laboratory, National Meirine Fisheries Service, 

 NOAA, Woods Hole, MA 02543. 



Manuscript accepted January 1988. 

 FISHERY BULLETIN: VOL. 86, NO. 2, 1988. 



counter rates for dolphins, and tuna catches. 



The maximization of profit is presumably a 

 strong influence on the behavior of commercial 

 fishermen. Upon leaving port, the catch rate real- 

 ized by a fisherman is probably the most impor- 

 tant factor affecting his profits. Given this orien- 

 tation, it is reasonable to assume that fishermen 

 have developed strategies of when and where to 

 fish that increase their catch rates beyond that 

 achieved by random search. There is little empir- 

 ical information to support this assumption other 

 than correlations, which have been noted for 

 some fisheries, between the spatial distribution of 

 catch rates and effort for the fleet as a whole (Gul- 

 land 1955; Calkins 1963). 



In a fishery where the detection of fish depends 

 upon visual cues, searching would be expected to 

 be located in the vicinity of previously located 

 fish, if there is a tendency for the underlying pop- 

 ulation to be spatially clustered. The search path 

 for a vessel in such a fishery might be expected to 

 look something like the hypothetical one depicted 

 in Figure 1. The amount of crisscrossing or 

 zigzagging in the vicinity of a catch and the area 

 over which the search extends would be expected 

 to vary between fishermen. The solution to the 

 optimal searching strategy for such a situation is 

 nontrivial and depends upon information on the 



351 



