144 



19 



To maximize effective population size, hatchery operators must: 



1. breed as many parents as is feasible; 



2. mate at least one male per female in daily matings; 



3. whenever possible, split gametes of the least numerous sex 



into subsets and cross each subset with gametes from a 

 different individual of the more numerous sex; and 



4. minimize variation in family size at the time of mating 



Recommendation 3 obviously places additional demands on labor and 

 other operational constraints of a particular hatchery. Before 

 implementing this mating procedure in a particular case, it may be 

 desirable to evaluate its operational costs and genetic benefit. Genetic 

 benefit, in terms of increased effective population size should be estimated 

 using equations for effective population size as a function of sex ratio 

 (Falconer 1981, Lande and Barrowclough 1987). Although variance in 

 progeny number is a very important variable that should also be 

 incorporated in this effective population size estimates (Lande and 

 Barrowclough 1987), the necessary data are currently unavailable for most 

 hatchery programs. This omission will not be corrected until families in 

 hatchery stocks are identified, as discussed below. 



Variation in family size can be the most important cause of reduced 

 effective population size and, thus, of erosion of genetic variation (Falconer 

 1981). Family size is the number of progeny, for one female or one male 

 parent, that survive until they reproduce themselves. Minimizing family 

 size variation in a hatchery stock requires mating similar numbers of 

 males and females from each family occurring in the entire group of 

 returning adults. This necessitates unique identification of each family 

 released from the hatchery so that surviving family members can be 

 recognized and counted when they return as spawning adults. One 

 approach would be to conduct incubation and early rearing in separate 

 compartments until fish are large enough to receive long-lasting tags (e.g., 

 PIT tags), and then to estimate family size by reading the tags of adults 

 returning to the hatchery. Future advances in application of DNA 

 fingerprinting methods to fish may eventually allow non-lethal 

 identification of families in the form of an intrinsic, biological tag. 



B. Avoid intentional and unintentional artificial selection in 

 collection of parents from donor stock and from adults 

 returning to the hatchery. 



