CYTOGENETICS AND EVOLUTION OF THE GRASS FAMILY l8l 



which are newer geologically than those of their diploid ancestors (Stebbins, 

 1950, pp. 348-350). Polyploidy, as a means of altering the genetic nature of 

 plants, is more likely to reduce than to improve their adaptation to the en- 

 vironment in which the original diploid species belongs and is well adapted. 

 This tendency has become clear from the results of many experiments with 

 artificially produced autopolyploids of various species of grasses and cereals, 

 all of which have turned out to be less well adapted to the original habitat 

 than their diploid progenitors. On the other hand, the radical change caused 

 by polyploidy can often promote the adaptation of the new types to entirely 

 different habitats from those occupied by their diploid ancestors. This is 

 particularly true when polyploidy is combined with hybridization, as usually 

 happens. The most common pattern of distribution and morphological varia- 

 bility which we find among genera of grasses and other plants containing 

 pol3^1oidy is consequently represented by the polyploid complex. Such com- 

 plexes typically contain two or more diploid species or subspecies, each with a 

 distinctive set of morphological characteristics and a well-defined range of 

 geographical and ecological distribution. As a rule, however, these diploids 

 represent only a small portion of the complex and are far outnumbered by 

 the polyploids. The latter may include a few types which are hardly distin- 

 guishable in outward appearance from their diploid ancestors, but a much 

 larger portion of the polyploids are easily recognized by their outward ap- 

 pearance. They are, however, usually devoid of distinctive morphological 

 characteristics not found in the diploids. On the other hand, they usually 

 contain various recombinations of the characteristics of the diploid species. 

 Furthermore, the polyploid members of such complexes are usually more 

 widespread in their geographic distribution than their diploid ancestors, and 

 instead of having neatly defined ranges of ecological tolerance often are 

 adapted to a great variety of habitats. Abundant evidence, some of which 

 will be presented below, shows that this wide range of ecological tolerance 

 possessed by natural polyploids has two causes. First, the polyploid repre- 

 sentatives of a complex include a greater variety of closely interrelated 

 genotypes than the diploids, and second, many of the individual genot5T3es 

 themselves can tolerate a wide range of different conditions. 



The polyploid complex thus emerges as a group of plants with a distinctive 

 evolutionary history, dominated by the processes of hybridization and chro- 

 mosome doubling. Successful polyploid complexes apparently arise from 

 recombination through hybridization between genetic types adapted to radi- 

 cally different environments or possessing different modes of adaptation to 

 the same environment. This genetic recombination builds up mechanisms of 

 adaptation to entirely new environments, to which none of the diploids are 

 adapted. Chromosome doubling may serve one or both of two purposes. 

 Species hybrids which are sterile because of differences in chromosomal pat- 

 terning of their parental species may be turned into fertile amphiploids, or 



