FISHERY BULLETIN: VOL. 87. NO. 4, 1989 



Figure 6. — Tracks of simulated purse seine vessels after 400 time steps during a simulation with a complex, static environmental 

 topography. Vessel movements are concentrated near peaks in the topogi-aphy, in i-esponse to the high density of dolphin schools 

 in these areas. 



characteristics of the environment produced 

 many more squares with relatively low densi- 

 tites. But here the process generating these 

 relatively low-density squares was very dif- 

 ferent from the simple environment case. In the 

 complex, moving environment the peaks were 

 moving at 1 knot. Aggregating the data from the 

 last 200 time steps generated a smeared version 

 of the underlying 16-peak array. Integrated over 

 the entire period of data collection, the areas of 

 dolphin concentration appear as bands across the 



simulated area, rather than as individual peaks 

 (Fig. 7). The dolphin schools spread themselves 

 out over a larger number of squares than in the 

 static case, producing lower estimates of aver- 

 age density per square. 



Stratification By Raw Encounter Rate 



Estimates derived from stratification by raw 

 encounter rate were relatively unbiased in all 

 cases because in these simulations we "collected" 



870 



