Hightower^ Rockfish harvesting policies in Washington-Oregon-California trawl fisheries 



651 



obtained. In some cases, the reduction was comparable 

 to the reduction that would be achieved by lowering 

 the CVs for biomass estimates from 50% to 25% or 

 from 25% to 0%. The percent reduction typically in- 

 creased as the CVs for biomassestimates increased, 

 although the loss in mean yield also increased some- 

 what (Hightower 1990). 



Single-species vs. multispecies policies 



When single-species (5) and multispecies (6) policies 

 were compared, the variance for total catch generally 

 was lower for the multispecies policy, except for some 

 cases for the five- species model or when biomass CVs 

 were 50%. In those cases, the better performance of 

 the single-species policy probably was due to the feed- 

 back mechanism it contained. For example, in the five- 

 species models, the single-species policies for shortbelly 

 and splitnose I'ockfish could adapt to the large changes 

 in the size of those stocks. When the five-species model 



was rerun using five "stocks" having chilipepper life- 

 history characteristics, the variance for total catch was 

 lower for the multispecies policy. 



For the single-species policy, the variance of total 

 catch typically was reduced by stabilizing the catch of 

 each individual species (Tables 3-5). For the multi- 

 species policy, reductions in the variance for total catch 

 were sometimes achieved by increasing the variance 

 for some of the less abundant individual species (Tables 

 3-5). The reduction in variance obtained using policy 

 (6) is consistent with Clark's (1984) prediction that a 

 multispecies system could be managed to reduce the 

 variance of total yield, relative to the variance of the 

 individual yields. 



For the multispecies policies, the optimal parameter 

 estimates were relatively insensitive to the objective 

 function and to the CV for biomass estimates (High- 

 tower 1990). The slope parameter bj i was less than 

 zero for all cases; consequently, fishing pressure varied 

 inversely with the alnmiiance of the other species. 



