304 K. STERN 



GENETIC CORRELATIONS IN RESISTANCE BREEDING 



Resistance is not what could be called a basic character (if there 

 are basic characters at all, except nucleotide sequences in DNA or sequences 

 of amino acids in polypeptides, etc.). It is a byproduct of some anatomical, 

 physiological, or biochemical feature of the host organism leading to 

 disturbance of a host-parasite relation (see Williams, 1964, for a review 

 for plant breeders) . This means that there must be genetic correlations 

 between resistance and other characters. These correlations can be 

 treated as any other genetic correlations, i.e., correlations of breeding 

 values (Falcolner, 1964) . But it seems very difficult to find solutions 

 for more complicated situations, because the theory of genetic correlations 

 is almost exclusively a theory of correlations between breeding values of 

 two or more characters, at least in its parts that are of some use to the 

 breeder. Naturally there must be other causes for genetic correlations 

 besides additive effects of genes on several characters. 



A breeder of long-lived organisms tests the resistance of his material 

 by laboratory tests or nursery experiments, and in field trials where 

 field resistance and yield of families or strains can be observed for long 

 periods. The more important genetic correlations will be detected easily 

 from comparisons of results of early tests for resistance and results 

 from long-term experiments. 



One of the most important genetic correlations is correlation of 

 resistance in different stages of development and/or in different environ- 

 ments. They can both be treated in the usual way, again if correlations 

 of breeding values are to be accounted for. The forest tree breeder 

 seems to be particularly interested in genetic correlation of seedling 

 resistance and resistance of mature trees. In fact, selection for white 

 pine blister rust resistance has widely been selection of resistant 

 seedlings where some degree of resistance during the complete life cycle 

 is the economic objective. Seedling selection or selection of young 

 plants from vegetatively propagated selected trees seems to be the best 

 approach. Haack (1914) inoculated mature trees when trying to evaluate 

 the heritable portion of resistance of Pinus silvestris L. against 

 Pevidermiwn pini in mature stands. A high proportion of his inoculations 

 were successful if the tree was already infected, but only few or none 

 were successful if the tree did not show spontaneous infections. This 

 method could serve as an excellent means for early evaluation of correla- 

 tions of resistance in seedlings and mature trees. It also gives some 

 indication of the degree of resistance since Haack found that the propor- 

 tion of successful inoculations in a tree was correlated with the number 

 of spontaneous infections in that tree. 



Mtllder (1952, 1953,, 1954, 1955) has made some comparisons of the 

 course of infection of a stand with increasing age. He found quite 

 different developments in stands of eastern white pine attacked by 

 Cvonavti-wn ribicola and stands of Scotch pine infected by Peridevmivm 

 pini , both fungi being closely related and very similar in reproductive 

 behavior. The average degree of resistance was much higher, of course, 

 in Scotch pine, the host-parasite system being much older here and having 

 probably reached some kind of equilibrium. But it is interesting to see 

 how the proportion of infected trees in both cases increases with age. 

 This is probably not only a consequence of a better chance for infection 

 during a longer period- -Mtllder found more infections in trees with larger 

 crowns and we cannot outrule this factor completely--but certainly also 

 a consequence of the threshold of resistance changing in time. It could 



