Relationship Between Genetic Diversity and Consistency of Performance in Different Environments 1

M UCH of the genetic improvement in self-fertilized crop plants has come through selecting "pure-line" varieties. The origin and reproductive biology of these varieties, plus the manner in which they are maintained by crop improvement associations and by seedsmen, cause them to be highly uniform in appearance--particularly as to size and maturity, resistance to diseases, and quality factors associated with market acceptance. Such varieties can also be very productive. Nevertheless, there is some evidence that mixtures may yield better, and more consistently, than stocks in which variability is kept to the minimum (6, 9, 14; review of older literature in Frankel, 4). Similar observations have been made by maize breeders who find that the highest yields in a particular environment are usually made by single crosses rather than by genetically variable double crosses. However, the same single cross rarely gives the highest yield in different years, or at different locations (13). The consequence is that average performance over a number of environments tends to be poorer for single crosses than for double crosses. According to Jones (10) the success of many double-cross hybrids over very wide areas is the result of their genetic diversity and its presumed consequence--stability and consistency of performance. Thus diversification is widely used in corn production although, as Jones pointed out, double-cross hybrids were originally adopted not for their consistency in production, but to overcome the handicaps that single crosses have in seed production. In self-pollinated crops, however, diversification has been little used since the unconscious cultivation of mixtures was brought to an end by the advent of pure-line varieties in the late nineteenth and early twentieth centuries. According to ~ensen (9), the concept of uniformity was stronger 1952 than ever before, On the other hand Frankel (5) believes that "a more functional attitude has grown up in recent years, with the recognition of the biological significance of ’purity,’ " and that ~the trend of development in breeding methods has been to provide a controlled measure of genetic heterogeneity." The theoretical advantages of diversity have been discussed repeatedly, and the rational blending of individual lines has been recommended periodically since the nineteenth century. Even so, there are surprisingly few comparisons between the productivity of genetically uniform and genetically diverse populations in self-pollinated crops. There is even less information on the role of genetic diverslty as a yield stabilizer. This paper presents evidence on these points from an experiment in which populations differing in level of genetic diversity were compared over a number of environments.