The Influence of Cytoplasmic Pollen Sterility on Gene Exchange between Populations.

YTOPLASMICALLY conditioned pollen sterility occurs rather frequently in plants. Usually, the character “pollen sterility” is produced when a particular cytoplasm is combined with a recessive gene in homozygous condition in the same plant. The best investigated example is the pollen sterility in corn (reviewed by DUVICK 1965), but similar conditions have been described in a large number of plant species. In nondomesticated plants, it is usual to find that the cytoplasm and the recessive genes which in combination with it produce pollen sterility occur in different populations. Cytoplasmic pollen sterility does. therefore, not usually occur in wild populations, but it appears when different geographical subspecies or closely related species are crossed. It has therefore been repeatedly suggested ( CASPARI 1948; JONES 195 1 ) that cytoplasmic pollen sterility may give rise to a mating barrier which would tend to inhibit gene exchange between different populations, and in this way initiate speciation. A mathematical model of cytoplasmically induced pollen sterility did not bear out this suggestion (WATSON and CASPARI 1960). This model demonstrates that if two populations interbreed which have two different cytoplasms and t-o pollen-fertility affecting alleles, such that the cytoplasm of one strain, combined with the allele from the other strain in homozygous condition induces reduced male fertility in its carriers, the following consequences will ensue: (1 ) The relative frequencies of the two cytoplasms will not be affected. (2) The genes contained in the two populations, including those affecting pollen fertility, will show a tendency to become equally distributed in the two cytoplasms. ( 3 ) The genes responsible for reduced pollen fertility in one of the cytoplasms will be selected against, and disappear from the mixed population. The result will therefore be a unified population, containing the two cytoplasms in their original frequencies (unless they have other effects on adaptive value) and homozygous for the alleles not producing pollen sterility in either cytoplasm. Recently, GRUN and AUBERTIN (1965), investigating crosses between different South American potato species, have found a number of cases of pollen sterility due to interaction between the maternally transmitted cytoplasm of one of the