GENETICS OF RUST FUNGI 11 



If the nectar drops of two different pycnia are mixed, and aecia 

 result from both, the two aecial progenies stem from reciprocal crosses. 

 The cytoplasmic contributions of spermatia to the aecial rudiment and 

 resultant dikaryon are likely to be very small. Differences between 

 aeciospores from reciprocal crosses will probably be due to cytoplasmic 

 inheritance. In this way cytoplasmic inheritance of virulence was demon- 

 strated by Johnson (1946, 1954) in crosses between different races of P. 

 graminis tvitici. 



Tests for homo- or heterozygosity of different loci are carried out 

 by selfing. This involves transfer of nectar among pycnia produced by 

 basidiospores derived from one dikaryotic clone. In such tests, Flor was 

 careful to make separate transfers between different individual pycnia, 

 maintaining the identity of their aecial progenies. Mi ah (1968) has dis- 

 cussed the relative merits of this method and others in which spermatia 

 of different pycnial drops are pooled and applied back to the same donor 

 pycnia. In the second method the progenies are derived either from aecia 

 or single aeciospores or from uredia or single uredospores , produced by 

 pooling the aeciospores. Sources of error include mixed pycnia which arise 

 from infection by two or more basidiospores of the same mating type, 

 cross-contamination of aeciospores, and the occurrence of multiple ferti- 

 lization of single aecial rudiments (Dinoor, Khair, and Fleischmann, 1968a) 

 Multiple infections may have a variety of effects on lesion phenotype , 

 some of which were described by Dinoor et at. (1968b) for ?. oovonata 

 avenae. Although tedious, single spore inoculations overcome most of 

 these problems. 



In the cereal rusts, the use of benzimidazole solutions (Person, 

 Samborski, and Forsyth, 1957) on which inoculated, detached leaves may be 

 floated in petri dishes has meant that rust free greenhouses, although 

 desirable, are no longer so important. Hooker and Yarwood (1966) cultured 

 P. sorghi through all stages on detached leaves of Oxalis oovniculata and 

 Zea mays, proceeding from uredospores to recombinant aeciospores in 5 to 

 6 weeks at any time of the year. 



Genetic analysis of blister rust on white pine would involve haploid 

 phenotypes generated by crossing two different haploid infections. Each 

 basidiospore could only be conveniently tested once because clonal propaga- 

 tion is not easy. Even so, it might be possible to find out if crosses 

 between rusts on two genetically different resistant hosts generate 

 basidiospores able to infect either host or the hybrid incorporating both 

 resistances. If foresters use vertical resistance to blister rust in 

 breeding white pines, then almost certainly such a demonstration will be 

 possible in the future. Some studies demonstrating meiotic recombination 

 in rusts are listed in Table 2. 



MITOT IC RECOMB IXAT ION 



The fact that rusts vary in the absence of meiotic recombination and 

 the discovery of the parasexual cycle in ksvevgillus prompted several 

 workers to look for heterokaryosis and parasexuality . Mixing uredospores 

 of two different rust clones is expected to generate two recombinant 

 dikaryons through exchange of partner nuclei, provided hyphal anastomosis 

 takes place. In fact such experiments do generate recombinants but often 

 many more than the two classes expected. For example Watson (1957b) and 

 Watson and Luig (1958) mixed red-spored race 111 with orange-spored race 

 NR-2 of ?. graminis tvitici. The orange race was virulent on 4 wheat 



