l80 STEBBINS 



to enrich the store of genetic material in an economically valuable species. 

 Furthermore, the physiologist who is seeking herbicidal chemicals which will 

 distinguish selectively between valuable grasses and noxious ones will be 

 helped much more by applying the newer system, with its emphasis on 

 physiological and cytological characteristics, than by judging relationships in 

 the traditional manner, solely on the basis of gross external morphology. 



THE ROLE OF POLYPLOIDY IN THE EVOLUTION OF GRASSES 



Recent studies of grass evolution have shown that the doubling of the 

 entire chromosome number, or polyploidy, has played an unusually large role 

 in the evolution of this family. Nearly all genera of grasses contain species 

 with chromosome numbers which are multiples of the original basic number. 

 In wheat and oats the various species contain 14, 28, and 42 chromosomes in 

 their body cells; in the panic grasses (Panicum) the numbers are 18, 36, 54, 

 and 72; while in the beard grasses (Andropogon) somatic numbers of 20, 

 40, 50, 60, 80, 120, and up to 180 have been found. 



If those species are considered pol3^1oid which possess multiples of chro- 

 mosome numbers which are known or can reasonably be inferred to exist in 

 their genera, then about 70 per cent of the species of grasses investigated 

 cytologically are polyploids. Since the percentage of polyploid species in the 

 flowering plants in general is about 30 to 35 per cent (Stebbins, 1950), 

 grasses have more than twice as many polyploids as the average for the 

 flowering plants. Furthermore, there is much evidence that the basic, or x, 

 number of many grass genera was itself derived by polyploidy in the remote 

 past, which suggests that the percentage of species with a record of poly- 

 ploidy somewhere in their evolutionary history is even greater. 



The geographic distribution of polyploid grasses does not favor either 

 the hypothesis of Tischler, that polyploidy is favored by cold climates, or 

 that of Hagerup, that dryness promotes polyploidy. To be sure, the fre- 

 quency of polyploids and the degree of polyploidy in many genera of the 

 Northern Hemisphere, such as Poa, Calamagrostis, and Bouteloua, become 

 higher as one goes northward. But this tendency can be explained on another 

 basis, namely, that the northern lands were not long ago occupied by a great 

 ice sheet which changed completely their topography and soil conditions. 

 Polyploids, being better adapted to new ecological niches, were better 

 equipped than their diploid relatives for the invasion of these lands as the 

 ice retreated. In respect to tolerance of drought the data are meager, since 

 few grass genera dominant in arid regions have been subjected to intensive 

 cytogenetic study. A survey of these data suggests, however, that polyploids 

 have been equally successful in both moist and arid regions. 



The most valid generalization which can be made about the relative dis- 

 tribution of diploids and polyploids is that polyploids usually occupy areas 



