GENETICS OF RUST FUNGI '13 



varieties to which the red race was avirulent. Only cultures from red 

 uredia on these 4 wheats were analyzed. Eleven different races stable on 

 further subculturing were found among 100 uredia tested. The information 

 on the genetic control of virulence to certain of the wheat varieties so 

 obtained agreed with information from selfing red-race 111 (Wilcoxson and 

 Paharia, 1958) . 



The generally accepted explanation for the appearance of this range 

 of classes is that, during growth on the selective host, nuclear fusion 

 and chromosomal reassortment occur to produce recombinant nuclei and hence 

 recombinant dikaryons. The observations of Dinoor &t at. (1968b) on the 

 range of phenotypes produced by multiple infections illustrates the impor- 

 tance of showing that recombinant phenotypes are in fact stable. This is 

 best done by single spore culturing. The appearance of such recombinants 

 during the vegetative growth of an isolated heterozygous dikaryon could 

 well explain much of what we have called spontaneous mutation. In the 

 rusts we can say little or nothing about the mechanism of meiotic or 

 mitotic recombination but there is little reason to doubt that it is 

 similar to Ustilago maydis and other fungi in which crossing over and 

 gene conversion can be recognized (Holliday, 1968) . While recombination 

 in diploids (of Ustilago maydis or Saocharomyaes eeveviseae) is now much 

 better understood, less attention has been paid to this in dikaryons. 

 Among the scarce published information on this subject are some intriguing 

 results of Ellingboe (1963) in the hymenomycete Schizophyllum commune 

 suggestive of genetic transfer of specific loci between separate nuclei. 

 Several examples of mitotic recombinations are noted in Table 2. 



The case of P. striiformis (Little and Manners, 1967) is of special 

 interest because no pycnial or aecial stages are known for this rust. In 

 this example the recombinant classes could be accounted for by reassort- 

 ment of intact nuclei. 



We should also note that recombination occurs between the varieties 

 of Puccinia graminis adapted to different cereals both by sexual hybridiza- 

 tion (Johnson, 1946) and by mitotic recombination (Bridgmon and Wilcoxson, 

 1959) . There seems to be no evidence in rusts of heterokaryon incom- 

 patibility systems like those found primarily in the ascomycetes and fungi 

 imperfecti which restrict heterokaryons to strains having a common 

 genetic background. 



DISCUSSION 



While technical difficulties have impeded genetic analysis of rusts 

 and linkage maps, biochemical genetics, and quantitative genetics have 

 still to be developed, the future holds many promises. The discovery 

 that certain Australian strains of P. graminis tvitiei can be cultured 

 on artificial media as dikaryons (Williams et at., 1966, 1967) opens the 

 way to work with induced auxotrophic mutants. These would be especially 

 useful if haploid mycelium can be cultured which, as far as I know, has 

 not yet been determined. 



Cultures also afford a much more convenient means of examining the 

 physiology of the rust organism away from its host . At the present time 

 mycelia are only available as spore germlings or fractionated by filtra- 

 tion and density gradient centrifugation , from macerated host tissue 

 (Dekhuijzen, Singh, and Staples, 1967). 



