on the biological characteristics of a population to make more 

 realistic assumptions about its growth rate, then MSY can be 

 determined more accurately. 



Assumptions underlying MSY calculations are that: 



1. the stock is self-contained; 



2. the stock was at carrying capacity before exploitation 

 began; 



3. there are no significant changes in carrying capacity 

 during the period of exploitation: 



4. the kind of density dependence of reproduction, growth 

 and/or natural mortality, and any time lags in the 

 stock's response to exploitation, will not cause fluc- 

 tuations of large amplitude in the stocks; and 



5. the process of reducing the 'initial' stock by exploi- 

 tation is reversible. 



These assumptions are very difficult to test and overly simplis- 

 tic (Holt and Talbot, 1978). 



Methods for calculating sustainable yields from a population 

 of a single species are well developed. The method used in a 

 particular case depends on which assumptions are applicable for 

 the patterns of recruitment, growth, natural mortality, fishing 

 mortality, and fishing effort (Ricker, 1975). 



Data for estimating rates of natural and fishing mortality 

 come from mark-recapture experiments, size or age composition of 

 fishery catch, or the relation of fishery success to effort (catch 

 per unit effort) . Estimates of recruitment or growth rate are de- 

 rived similarly, and may have additional information from life 

 history or physiology of the species, e.g., how many eggs a gravid 

 female produces. It is important to note that most of the data on 

 which fishery yield calculations are based come from fishing. When 

 exploitation of a population stops, there is no longer as much 

 information on the condition of the stock. 



If data are not sufficient to base MSY calculations on vari- 

 ous population models with well defined assumptions about growth 

 and mortality rates, then sustainable yield levels are estimated 

 through a process of educated guessing and experience. Yield es- 

 timates can be based on the apparent effects, or lack of them, 

 on a stock of recent or historic catches. This is dangerous since 

 "effects" may not be apparent if substantial time lags are in- 

 volved. Yield estimates might also be based on information about 



6-3 



