THEORETICAL BASIS OF RUST RESISTANCE TESTING 303 



schemes like reciprocal recurrent selection are to be preferred in such 

 cases. 



There seems to be hardly a chance for epistatic gene effects to play 

 a major role in equilibrium systems of host and parasite, because 

 epistatic equilibria are possible only in exceptional cases. But the 

 reverse situation can be given if the parasite has been newly introduced 

 or breeding for resistance in a territory that has been newly opened for 

 the host. Both situations in principle are similar to what in population 

 genetics has' been circumscribed by the terms, "use of preadapted geno- 

 types" or "use of recruited genotypes" (see Spiess , 1968, for summary 

 and references) . 



Recruitment of genotypes means the emergence of hitherto unknown 

 qualities of some--often very rare- -genotypes of a population after the 

 population has been exposed to a rather drastic environmental stress. 

 Preadaptation means that some of these genotypes are better adapted to 

 the new environment; they might, for instance, be more resistance to a 

 newly introduced disease. It is therefore the special case of recruit- 

 ment where useful genotypes have been recruited by nature or by the 

 breeder. Epistatic gene effects might be one of the major causes of 

 preadaption, as pointed out by Wright (1956) . Crosses between preadapted 

 individuals thus need not necessarily result in resistant progenies. 

 But it should be possible to find parents of high specific combining 

 ability among the group of preadapted genotypes or to develop appropriate 

 crossing designs for the use of this type of genetic variation. 



Bingham et at. (1969) could show that the proportion of nonadditive 

 genetic variance of the total genetic variance increases with intensity 

 of infection (proportion of infected individuals) . This is in full 

 agreement with our expectation. As long as the infection rate is low, 

 small differences in phenology might be of some value in increasing 

 resistance. Under normal circumstances this means possessing characters 

 adaptive to environmental factors rather than resistance to a newly 

 introduced parasite. The genetic variances of such characters are pre- 

 dominantly of the additive type. But other preadapted types must be 

 "recruited" from the population if the probability of infection increases 

 Under such conditions the only truly resistant genotypes might be those 

 which normally have inferior selective values or might even belong to 

 the genetic load of the host population. This is a troublesome case for 

 the breeder since epistatic genetic variance is difficult to assess in 

 the common type of experiments like the diallel cross. Kearsey and 

 Jinks (1968) have given some new ideas on how to proceed in such cases. 

 But much more work should be done or must be done before breeders are 

 able to make use of epistatic genetic variance in a sufficiently simple 

 way. 



Vogl, Schonbach,and Haedicke (1968) give an example of preadaptation 

 in resistance to an abiotic damage. Different provenances of Japanese 

 larch proved to be of different resistance to air pollution, the degree 

 of resistance being correlated to some unknown or only partly known 

 adaptation to the climate in the places of origin. This is probably an 

 example of preadaption of the type mentioned first. 



