POLYPLOIDY 



SOME GENERALIZATIONS ON POLYPLOIDY 



It appears, then, that allopolyploidy is a very general phenomenon. In 

 1942, Goodspeed and Bradley published a review in which they listed 

 124 well-authenticated cases of allopolyploidy, including both natural and 

 experimental examples. There is no doubt that the list would be much 

 longer now if it were brought up to date. Natural polyploids seem to be 

 generally allopolyploids rather than autopolyploids, and so this type evi- 

 dently has especial interest for evolution. As pointed out, inspection of 

 tables of chromosome numbers makes it appear probable that over half 

 of the higher plants are polyploids of one degree or another. While the 

 formation of polyploid series does not entail any new genie material, it 

 does produce new combinations upon which selection can act, combina- 

 tions which may be very different from anything formed in any other way. 

 It is not surprising, therefore, that polyploids often invade territories not 

 occupied by their diploid parents. Polyploids seem to be much more 

 aggressive invaders of new territory than are their diploid relatives. For 

 example, Anderson has shown that Iris versicolor is almost certainly an 

 allopolyploid derived from I. virginica and Z. setosa. I. virginica is widely 

 distributed in the southeastern portion of the United States. I. seiosa is 

 found as two widely separated races, one on the coast of Alaska, the other 

 in Labrador, Nova Scotia, and Newfoundland. These species are regarded 

 as remnants of the preglacial floras of their respective areas, and they have 

 not extended their ranges into the glaciated parts of North America. But 

 I. versicolor, their allopolyploid offspring, is distributed from Labrador 

 and northeastern United States westward through the Great Lakes region 

 to Wisconsin and Winnipeg. It is thus widely spread in the glaciated part 

 of North America. 



An interesting and largely unsolved problem is that of the effect of 

 polyploidy on the occurrence of new genetic variability due to mutations. 

 Some geneticists have expressed the opinion that polyploidy should in- 

 crease the total genetic variability rapidly, because random mutations 

 might occur in any of the genomes and become homozygous. Others have 

 pointed out that a new recessive mutant occurring in one genome would 

 now be covered by its dominant allele in tJiree other genomes ( or more in 

 higher polyploids) so that its phenotypic expression would be much less 

 probable than in a diploid. It may be that these viewpoints are not com- 

 pletely irreconcilable. From a short-range viewpoint, the latter is prob- 

 ably correct, but, when time is provided on a geological scale, differentia- 

 tion of the genomes by random mutation should finally result in greater 

 total variability than could be achieved in a diploid. 



It appears to be generally true that diploids are found in the older part 

 of the total range of a group, while polyploids invade the geologically 

 more recent parts. Thus, in a group in which polyploidy is common, the 

 diploids tend to become relics, while the more diversified polyploids fill 

 the available niches. Stebbins pointed out that one consequence of this 

 is that, when conditions become unfavorable for such a group, the diploids 

 are the first to become extinct. And, when a mere relic remains of a once 



305 



