240 



Fishery Bulletin 100(2) 



catch estimates for the season-permit anglers C"' were the 

 same value for metho(i 1 and method 2. The population 

 size N and the sample size /; were in proportion constant 

 throughout the season; therefore NJn^ = NIn was same for 

 all strata. The ratio of the variances is 



V(Y^ ) {N,+ N., fU'fViy, ) + nlV(y, )| ' 



which also shows that the precision of both methods 

 depends on the sizes of the strata, the sample sizes, and 

 the variance of the sample means. If the allocation is 

 proportional, the variance ratio becomes 1 and the two 

 variances coincide. The objective of stratified surveys is 

 to obtain precise total effort or catch estimates (or both) 



over the entire season. A proportional sample allocation 

 is recommended, which allows a simple calculation with 

 method 2 and improves precision of estimates at the same 

 time. 



Acknowledgements 



We extend our gratitude to the anonymous referees and to 

 John V. Merriner, Katherine Myers, and Sharyn Matriotti 

 for their critical readings that greatly improved the man- 

 uscript. We also thank John M. Hoenig, John B. Pearce, 

 Yashushi Taga, and Geoff Gordon for their comments on 

 earlier versions of the manuscript. A portion of this study 

 was funded by the Fisheries Agency of Japan (Japan Sea- 

 Farming Association ). 



