An article in the current issue of the Journal of Heredity (a publication of the American Genetic Association) is of potential interest to killifish breeders: Frankel, J.S. 1997. Inheritance of body coloration in the lyretail toothcarp (Aphyosemion australe Cyprinodontidae). J. Hered. 88(5): 445-446
Frankel shows that the inheritance of the orange body color trait is determined by two independently assorting autosomal (non sex-linked) genes with recessive expression. Homozygotes (genotype aabb) are orange. So are fish with genotypes aaB__ and A__bb. All other genotypes (A __ B__) are brown ("wildtype").
Thus, one could make double heterozygotes by crossing pure-breeding browns (genotype AABB)with pure-breeding oranges (aabb), and then intercross the heterozygotes and obtain a ratio of 9 brown to 7 orange (roughly 44% orange), a more favorable result than the 3:1 ratio of brown to orange one would have if the orange trait was encoded by a single recessive allele as, for example, albinism in many species. Backcrossing of the F1 hybrids to orange individuals would yield a 3:1 ratio of orange to brown, even more favorable
In my opinion these results are of increased significance because the orange (= "red orange," = "golden") strain of A.australe is very popular, particularly among beginning killifish hobbyists, and it enjoys steady commercial interest as well. Although I don't have a lot of data on this, it is my impression that in recent years our stocks of the orange variant have become increasingly difficult to breed successfully. In my current stock, for example, I frequently get poor hatches unless I incubate the eggs on a peat layer rather than leave them in water. Though fertility does increase with age, many of my young pairs are almost sterile. This fish was so easy to breed in the early 1960's that I probably would not have been able to believe that problems breeding it might eventually develop. Presumably, these problems are due to inbreeding depression (though in my case they may be complicated by differences in water quality).
The time-honored way of dealing with inbreeding depression in a "cultivar "such as the orange lyretail would be to cross it back into authentic wild stock and then intercross the F1 progeny or backcross them to the cultivar, picking up much of the genetic variance that might be present in the wild breeders. Frankel's article shows that this could be a rather efficient and practical enterprise with the orange A. australe, for these matings could yield good proportions of orange progeny.
One word of caution though: if one were to do this, it should be done with stock collected at the same locality ("Cape Lopez") as the original stock which gave rise to the orange cultivar or as close to that locale as possible. Many Aphyosemion "species" are comprised of populations which can appear to be very similar in color pattern but which are really are multiply divergent in chromosome number and/or shape; this certainly includes the A. ahli complex, of which A. australe is a member. Thus, by crossing the "wrong" populations one could inadvertently produce heterozygotes for different chromsomal arrays, and these are quite frequently sterile (read "mules") or of reduced fertility. This we do not need. It would simply not do to "solve" one genetic problem by introducing another!