The fecundities of 55 genotypes of the form SD(i)/SD(j) generated by 11 different SD chromosomes have been examined. Five of the genotypes are lethal The fecundities of the rest fall into a pattern of fertility and sterility that is highly suggestive of intracistronic complementation. The complementation leading to male fertility is only partial complementation: the fecundity of most fertile genotypes is less than half that of controls. The three components of the SD system, the Sd locus, the Ac locus, and the modifiers in 2R, were examined separately, and it appears that the complementation is a phenomenon associated with the Sd locus. A hypothesis of the molecular events involved in segregation distortion is formulated in the light of these observations. The model is based on the assumption that the Sd locus produces a multimeric molecule that regulates the activity of the Ac(=Rsp) locus during spermatogenesis.
The recovery of the SD chromosome from a heterozygous SD male increases with brood. This is independent of the age of the female, occurs during the time the sperm are stored in the females, disappears when the segregation distortion is suppressed, and is temperature-sensitive-temperature shocks above or below 25 degrees C applied to the mature sperm both tend to accelerate the increase in the recovery of SD. All this suggests the existence of a class of sperm affected by SD in which the sperm are able to fertilize eggs for a short time following ejaculation but become dysfunctional thereafter.
A model of segregation distortion is assumed in which the action of the distorter when heterozygous is to render dysfunctional those gametes that carry its allele. Two gamete killers when homozygous are assumed to distort each other. Individuals that carry the gamete killer suffer a reduction in the number of functional gametes they produce, but this deleterious effect is counterbalanced by the segregation ratio advantage of the distorter. The dynamics of such a system are analyzed in terms of a generalized fecundity function, which is defined as a function which assigns to any individual his relative fecundity in terms of the fraction of functional gametes he produces. Three general classes of fecundity functions are considered: (a) proportionality, in which the relative fecundity of an individual is proportional to the fraction of functional gametes he produces, (b) compensation, in which the relative fecundity of an individual is always greater than the fraction of functional gametes he produces, and (c) mass action, in which the relative fecundity of an individual is less than or greater than the fraction of functional gametes he produces according to whether the fraction of functional gametes is less than or greater than some threshold. In case (a) all gamete killers are always at neutral equilibria and gene frequency changes at the locus are governed by random drift. In case (b) all gamete killers will be fixed if the fecundity function is such that its second derivative is negative, whenever its argument is greater than one-half. And in case (c) some gamete killers will converge to stable equilibria, others will be fixed. If a gamete killer is homozygous lethal it will almost always converge to a stable equilibrium.
Drosophila males heterozygous for the segregation distorter chromosome show a reduction in fecundity which is correlated with the degree of distortion of their segregation ratio. When males are made homozygous for the segregation distorter elements, their fecundity is lowered to just that extent expected if each SD were independently causing the sperms carrying the other to dysfunction. In three cases tested, Sd Ac/Ac, Sd Ac/Sd Ac, and Sd Ac St/Sd Ac St, the fecundity of the males is consistent with the dysfunction model, whereas the females are normal. The Sd Ac St/Sd Ac St males are almost completely sterile. They do produce sperms, however, which are motile and which look morphologically normal in the phase contrast microscope. They are transferred to the female during copulation and are stored in apparently normal numbers in the seminal receptacle and spermathecae.