While this may be less of an issue with fish, interspecies mixing is the best way to weed out species specific weaknesses. Pure bred is all well and good, but the idea that species purity gives the animal an advantage is crap.
I think there's a mixup between species and breed here.
For a species to succeed in the wild, it has to be strong enough to do so. In general, hybridization leads to sterile, weak progeny that are quickly removed from the gene pool. Or, as Alcock says in
Animal Behavior , "Hybrid offspring are likely to develop poorly, if at all, and even if they succeed in reaching maturity, they usually have scrambled adaptations that match neither parental niche and leave the hybrids at a disadvantage in the competition for resources." That's why there are behavioral mechanisms for species separation in the wild. In regions where closely related species overlap, one often sees these mechanisms amplified, such as increased differences in the frequencies of mating songs. There are naturally occurring hybrids, such as the "red wolf" or some baboons, but these are the exception, rather than the rule.
On the other hand, dog breeds are all part of the same species. Dachsunds and great danes are all Canis familiaris. Many breeds have such small gene pools that deleterious mutations are exposed at a very high frequency. Same thing would happen if brother and sister marry. It's also why zoos keep registries of rare species, to optimize genetic diversity. And why mutts tend to be more vigorous.
As far as fish go, I just want to know what I have. Once species hybridize, it's impossible to unmix the DNA, so subsequent generations will be a genetic gamisch. Plus, it makes it hard to predict the appearance of the progeny.
This is feeling a lot like GCC.