SPECIFICITY IN PLANT DISEASE 35 



not have the corresponding genotype for high pathogenicity (Loegering and 

 Powers, 1962). Flor used Bison flax as his "universal suscept" (Person, 

 1959) . In Australia Kerr (1960) found cultures of M. lini which are 

 avirulent on this variety, indicating that Bison does have at least one 

 gene for low reaction and the population of M. lini in the United States 

 is homogeneous for the corresponding gene for high pathogenicity. 



All accumulated evidence related to the gene-for-gene hypothesis has 

 been obtained from host : pathogen relationships in which the vegetative or 

 repeating stage of the fungus has been studied, whereas the sexual part 

 of the life cycle of these pathogens is not significantly involved in 

 producing damage to the economic host. In the case of blister rust, how- 

 ever, it is the pycnial and aecial stages (non-repeating stages) of C. 

 vibicola with which we are concerned. The work of Flor involved a fungus 

 which is autoecious- -having all stages of the life cycle on one host 

 species--but the vegetative (uredial) stage, which is the damaging stage, 

 was studied. One study (Flor, 1959a) of the pycnial and aecial stages is 

 of considerable interest. Flor induced sporidial infection from telio- 

 spores of a culture of race 201 of M. lini on the Bison and Bombay varieties 

 of flax. Using the resulting pycnia, he made 5 selfs on Bison and 58 on 

 Bombay from which he obtained aecia. All 43 cultures were virulent on 

 Bison but all were avirulent on Bombay. Thus Bombay behaved like an aecial 

 host. These results could be obtained as a result of a single gene dif- 

 ference between Bombay and Bison. 



Puccinia gvaminis tvitici is a heteroecious fungus, and it is the 

 uredial stage which occurs on the economic host (wheat) . It is this stage 

 in which the extensive specialization is found as a result of numerous 

 corresponding gene pairs for reaction and pathogenicity. In general its 

 pycnial host, barberry {Bevbevis spp.) 3 is "susceptible" just as white 

 pine is "susceptible" to C. ribieola. Some cultures of P. gvaminis tvitici 

 have been shown to be avirulent to barberry (Green and Johnson, 1958; 

 Johnson and Green, 1954), which indicates that there might be a gene-for- 

 gene relationship between barberry and P. gvaminis tvitici just as there 

 is between wheat and P. gvaminis tvitici. 



The relationship between host and pathogen sometimes becomes confused 

 by the taxonomic classification of the host and/or the pathogen. The two 

 blister rust fungi, C. vibicola and C. occidentale Hedge. , Bethel & Hunt, 

 might be considered the same species since both alternate to Ribes spp . 

 There are some morphological differences between them, but they are 

 differentiated mainly by the fact that their pycnial-aecial hosts are 

 different pine species. The two Peridermium rust fungi, C. fusifovme 

 Hedge. £ Hunt ex Cumm. and C. quevcuurn (Berk) Miyabe ex Shirai, can be 

 differentiated by the reaction of Quevcus velutina Lam. to the uredial 

 stage of these two fungi (Dwinell, 1969). Low infection type results from 

 C. fusifovme: Q. velutina and high infection type results from C. quevcuwn: 

 Q. velutina. However, C. quevcuurn can be divided into two physiologic 

 races by the differential reaction of Pinus banksiana Lamb, and P. 

 vivginiana Mill, to infection by sporidia (Powers, in press). 



In leaf rust of wheat d'Oliveira (1966) has shown that in Portugal 

 Puccinia vecondita Rob. ex Desm. f.sp. tvitici can be divided into two 

 groups by the differential reaction of the aecial hosts, Thalictvurr, 

 speciosissimum Loefl. and Anchusa italica Retz. Cross fertilization of 

 pycnia on these two hosts has not been possible. This specialization on 

 the aecial hosts does not appear to be related to the specialization of 

 the uredial stage of the fungus on wheat. If we compare leaf rust of 



