56 J. C. ZADOKS 



therefore, less suitable as a criterium. The sporulation rate is easily 

 measured and it may be a valuable criterium. The epidemiologist considers 

 these components of resistance, especially infection ratio, latent period 

 and sporulation rate, as suitable selection criteria. Information on 

 the heritability of these components of resistance is, unfortunately, 

 scarce . 



EPIDEMIOLOGICAL CONSEQUENCE OF BREEDING TACTICS 



Breeding tactics have epidemiological consequences. There is evi- 

 dence that the potato cultivars used in Europe around 1845, when 

 Phytophthora infestans first spread over the continent, were much more 

 susceptible than the cultivars grown later, in the intervening period 

 before the introduction of differential resistance. In the interim 

 farmers reaped the fruits of natural selection for partial resistance, 

 now thought to be uniform (van der Plank, 1963). 



A more recent example comes from maize. Tropical maize rust, 

 Puooinia polysora, was first introduced into Africa in the late 1940 's. 

 After a few years of severe losses the local maize populations recovered 

 and tropical maize rust is no longer a major threat. Again, natural 

 selection produced an adequate partial resistance, probably uniform and 

 of polygenic inheritance (van der Plank, 1968). 



In the corn belt of the U.S., the other maize rust, P. sorghi, is 

 more prevalent. Many races of this rust are known but nevertheless 

 known genes for differential resistance are not commonly used in breeding 

 programs. The partial but uniform resistance obtained without much 

 conscious effort is satisfactory. 



The story of differential resistance is in dramatic contrast to the tale 

 of uniform resistance. Differential resistance was first used around 

 1920 to control diseases like stem rust of wheat (P. gramtnis) and late 

 blight of potatoes (P. infestans) . Unfortunately, the pathogens showed 

 a remarkable adaptability. Disease-free crops carrying the new resis- 

 tance genes provided an ideal medium for the selection of genes for 

 differential virulence in the pathogens. The result was a race between 

 breeder and pathogen, the breeder producing new cultivars carrying new 

 differential genes for resistance and the pathogen adapting itself time 

 and again by producing the compatible genes for differential virulence. 

 The average useful life of wheat cultivars in Mexico is about five years 

 because of stem rust (Borlaug, 1965). The same is approximately true in 

 the Netherlands where the appearance of new stripe rust races (P. 

 striifovmis) in commercial wheat fields has a frequency of nearly one 

 per year. 



Most of the differential resistance genes used are dominant and 

 therefore epistatic to the polygenes conditioning uniform resistance. 

 Effects of the polygenes for resistance are masked, and as a consequence 

 they are diluted in successive breeding cycles until uniform resistance 

 is lost. This phenomenon is known as the "Vertifolia effect" (van der 

 Plank, 1963). The Vertifolia effect is dangerous because the uniform 

 resistance can serve as a second line of defense when the first line of 

 defense, differential resistance, has been broken. In the pathogen, 

 differential genes for virulence characterize the physiological races as 

 the breeder knows them. They are produced by mutation, they are combined 

 and recombined by sexual or parasexual processes, and they are selected 



