I think that would be possible if I were keeping all individuals from each year class in the program. However since I always only take the fastest growing fish I believe that I'm keeping any defective genes from entering the population. If I get a defective recessive gene it would show up in 25% of any subsequent generation and would be weeded out at the next selection level when I pick the top 5 percent. Now if I get a defective dominant gene, then I've got a little bigger problem and would need to discard a year class and re-pick different individuals from the previous generation and try again. I'll lose a full year if this happens. It's a pretty good fail-safe system. So far each generation's select fish have grown faster than their previous generation counterparts. As long as this trend continues I don't think I have anything to fear. I have quite a few age-4 and 5 fish from the F-2 and F-3 generation and they still are showing good fitness and growth. You saw some of those fish in the seine this spring.
