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



SCHOOL DENSITY : EST . vs ACTUAL 



8000 



SIMPLE 



GENTLE 



(2,1) 



COMPLEX 



STEEP 



(4,5) 



>- 

 CO 



o 



O 



X 



o 



CO 



Q 

 lU 

 I- 

 < 



CO 

 lU 



High Density 



STATIC DYNAMIC 



2500 



8000 



STATIC 



DYNAMIC 



ABCABCABCABC 



1250. — +--— 



ABCABCABCABC 



DATA TREATMENT 

 A Not Stratified ; Not Smoothed 

 B Stratified ; Not Smoothed 



C Stratified ; Smoothed 



Figure 4. — Estimated vs. actual school density (total number of schools in 

 simulated area) under 8 different cases for model conditions and under 3 types 

 of data simulation for each condition. Each set of 4 columns is a set of 4 

 replicated runs for a given case. The cases differed in 1) environmental to- 

 pogi-aphy (simple, gentle vs. complex, steep), 2) whether the topography was 

 static or dynamic (sliding left at 1 knot), and 3) actual abundance of dolphin 

 schools (1,250 or 2,500). Numbers in parentheses (2, 1; 4, 5) are parameters 

 used in equations generating the topogi-aphies. Two and 4 refer to the number 

 of peaks arra.ved along each axis of the spatial plane, generating a regular 

 square grid of peaks. One and 5 are values of the parameter controlling peak 

 slope: 1 generates a gradual slope, 5 generates a precipitous slojje. Heavy 

 lines across the figures indicate the true density (abundance) of schools in each 

 set of simulations. Data treatments include A) no stratification before estimat- 

 ing school abundance, B) stratification of 1" squares based on observed (raw) 

 encounter rates per square, and C) stratification of 1° squares based on 

 smoothed encounter rates per square. 



dant, and avoided areas where schools were few 

 (Fig. 6). Overestimates of average school abun- 

 dance per r square resulted from this pattern of 

 effort because few samples from low density 

 squares contributed to the average. Overesti- 



mates of total school abundance then followed 

 directly by extrapolating this overestimate to 

 the entire area. 



Overestimation of school abundance was espe- 

 cially pronounced for the case of a complex, 



868 



