476 



Fishery Bulletin 103(3) 



catches in the LFD area was higher than in the HFD 

 area. For Pacific halibut, the average proportion of zero 

 catches remained approximately constant across the 

 entire habitat area. The relative mean nonzero catch 

 between the LFD and HFD areas varied across species, 

 ranging from 37% to 82% (Table 4). 



In each of the 24 species-year combinations, three esti- 

 mates of population abundance were compared, except for 

 flathead sole in 1992 when no samples were taken in the 

 flathead sole HFD area (Fig. 4). In every case in which 

 the proportion of habitat stratum-size sites to total study 

 area sites exceeded the proportion of habitat stratum 

 size to total study area size (Table 2), the unstratified 

 estimate was greater than the estimate poststratified by 

 habitat (Fig. 4). In every case that the proportion of habi- 

 tat stratum sites to total study area sites was less than 

 the proportion of habitat stratum size to total study area, 



the unstratified estimate was less than the estimate 

 poststratified by habitat. Similarly, in every case that 

 the proportion of HFD stratum sites to habitat stratum 

 sites exceeded the proportion of HFD stratum size to 

 habitat stratum size (Table 2), the estimate poststratified 

 by habitat was greater than the estimate poststratified 

 by habitat and fish density (Fig. 4). In all but two cases 

 in which the proportion of HFD stratum sites to habitat 

 stratum sites was less than the proportion of HFD stra- 

 tum size to habitat stratum size, the estimate poststrati- 

 fied by habitat was less than the estimate poststratified 

 by habitat and fish density. The two exceptions were for 

 Pacific halibut in 1991 and 1996, where the difference 

 between poststratified estimates was small. In 1991, the 

 estimate poststratified by habitat was 2.9% (8116 fish) 

 greater than the estimate poststratified by habitat and 

 fish density; in 1996, it was 0.56% (4905 fish greater). 



