g vs. 3,439.2 g) when F-multipliers were large, 

 which is impossible. Maximum biomass of the 

 stock was estimated at F equals zero; the time 

 intervals used were four one-eighth of a year 

 intervals followed by one-half year intervals. De- 

 spite the small intervals, the difference between 

 the yield per recruit estimates was large, indicat- 

 ing a need to minimize the G t - Z t difference if B t 

 is calculated arithmetically, regardless of the size 

 of the time intervals. Therefore, in similar cir- 

 cumstances and for the example data set, we rec- 

 ommend that B t be calculated exponentially. 



Acknowledgments 



We thank David R. Colby, William E. Schaaf, 

 Dean W Ahrenholz, and William R. Nicholson of 

 the Southeast Fisheries Center Beaufort Labora- 

 tory for their discussions and review of the manu- 

 script. Figures are by Herbert R. Gordy 



Literature Cited 



PAULIK, G. J., AND W. H. BAYLIFF. 



1967. A generalized computer program for the Ricker 

 model of equilibrium yield per recruitment. J. Fish. Res. 

 Board Can. 24:249-259. 

 RICKER. W. E. 



1975. Computations and interpretations of biological 

 statistics of fish populations. Fish. Res. Board Can., 

 Bull. 191, 382 p. 



SHERYAN P EPPERLY 



Department of Marine Science 



University of South Florida 



St Petersburg, FL 33701 



Present address: 



North Carolina Division of Marine Fisheries 



P.O. Box 769 



Morehead City, NC 28557 



WALTER R. NELSON 



Southeast Fisheries Center Beaufort Laboratory 



National Marine Fisheries Service. NOAA 



Beaufort. NC 28516 



Present address: 



Southeast Fisheries Center Mississippi Laboratories 



National Marine Fisheries Service. NOAA 



P.O. Drawer 1207 



Pascagoula, MS 39567 



448 



