RACE DIFFERENTIATION IN CEREAL RUSTS 67 



Manners (1950) in his yellow rust wotk came to the same conclusions 

 pronouncing "Stakman' s rule" (Stakman, 1947) that a new "physiologic race 

 of rust must not be established unless there is a consistent difference 

 in reaction type on one or more hosts at least as great as that between 

 susceptible and mesothetic, or between mesothetic and resistant." The 

 list of races of the other cereal rusts, which have an alternate host 

 and which are far more investigated, are much longer, like the list of 

 stem rust races of wheat with 297 races in 1962 (Stakman, Stewart, and 

 Loegering, 1962) . 



New epidemics on hitherto resistant varieties very often create a 

 new differential. For example, the varieties Cleo, Falco, and Opal were 

 not infected with race 3/55 until 1960 and have remained susceptible 

 (Ubels, Stubbs, and s 'Jacob, 1965). Unfortunately Cleo and Falco are 

 not seedling differentials, since reactions on seedlings do not distin- 

 guish between and IV infection types. The same is true for the 

 famous Heine VII, which has a rather distinguished stripe rust response 

 in the field but not in the greenhouse in seedling stage plants (Fuchs, 

 1960) . The variety Opal seems to be a much better possibility of a new 

 differential variety both in the field and in the house. 



Another example of the creation of a new differential and, at the 

 same time, a new race happened in the United Kingdom and the Netherlands 

 in 1965 (Macer and Doling, 1966; Fuchs, 1967), and gives another illus- 

 tration for the race between breeder and fungus in nature. The variety 

 Rothwell Perdix had shown excellent resistance against the main northern 

 European races 3/55, 8, 27/53, and 54. When planted to a larger extent, 

 however, it showed a stripe rust infection that has increased every year 

 since then (only last year it seemed to stop spreading to other countries) 



In Table 1 you will find the main European stripe rust races charac- 

 terized by their behaviour on some differential wheat varieties, both in 

 the field and on the seedling stage in the greenhouse. Michigan Amber 

 is the generally susceptible host, highly infected with all races. 

 Vilmorin 23 and Cappelle are susceptible to the 3/55 race but resistant 

 to all other races mentioned here. Chinese 166 is susceptible to 27/53 

 and 60 races, and Heines Kolben is susceptible to 54 and 60 races. 

 Rothwell Perdix was resistant to all the older races, but was susceptible 

 to the new race 60. Heine VII exhibited uniform susceptibility in the 

 seedling stage in the greenhouse, but showed differential resistance 

 when mature plants were tested in the field. 



If Opal were brought into this system, it would show its "additional 

 differential" character, allowing a subdivision in the race 3/55 group, 

 while Rothwell Perdix has an independent "main differential" character. 



People working with differentials of variable reliability are always 

 happy to meet good additional or supporting differentials. I started to 

 search for a good additional differential myself, by testing in the 

 greenhouse hundreds of varieties that promised some hope for this purpose. 

 Four groups were defined as follows: 



(1) generally susceptible (the main amount) , 



(2) differentiating to some extent, but not very reliable (rather 

 small amount) , 



