THEORETICAL BASIS OF RUST RESISTANCE TESTING 309 



Australia where wheat is a relatively new crop--anyway where wheat hasn't 

 been subjected to stripe rust for the last 100 years. Here there should 

 be neither selection for or against resistance genes. They should float 

 in populations according to the Hardy-Weinberg principle. White pine 

 trees (assuming they haven't been exposed to blister rust for a few 

 thousand years) should contain some similar genes. Of course resistant 

 gene-complexes that provide spectacular resistance might be broken up 

 into single units, imparting little or no resistance separately. I have 

 not watched Bingham's western white pine blister rust resistance work too 

 carefully, nevertheless everything he says points toward multiple-gene 

 resistance. He has some resistant mature trees that produce progeny 

 seedlings only a few of which are resistant; other resistant parents 

 produce more. This suggests multiple-gene heterozygous resistance, and 

 a difference between seedling and mature plant resistance. If the blister 

 rust situation matches that in stripe rust, the main effect is from the 

 total number of genes; that is gene dosage, not dominance. I don't know 

 whether the severe, screening-inoculation of the young white pine seed- 

 lings is comparable with the situation in wheat. In my own work wheat 

 plants with good mature plant reistance can be seedling susceptible. Thus 

 I have avoided known seedling resistant wheats on the assumption that 

 resistance was controlled by specific genes that would obscure the more 

 non-specific resistance I was watching. I would suggest that you spend 

 more time looking at trees instead of reviewing the literature and 

 worrying about its implications in trees. Your tree^ haven't read this 

 literature, and yet you do have wonderful, fairly old, blister-rust- 

 resistant trees out there in the forest. From what you tell me, unless 

 these trees produce a high proportion of resistant seedlings you don't 

 use them. I think this is a colossal mistake. You need more than seed- 

 ling resistance. Use some trees that have seedling resistance but put 

 all of these trees together in a way they "want" to go together, instead 

 of how you decide they should go together. If you let me warm up I'll 

 add some more commentary on another day. 



STERN": This has been precisely what I was referring to--to the two 

 combination effects "general combining ability", contributing to the 

 additive genetic variance, and "specific combining ability" contributing 

 to the non-additive genetic variance, as outlined by Sewall Wright (1956). 

 Wright gave an example (after King, J., 1955. Integration of the gene 

 pool as demonstrated by resistance to DDT. Amer. Natur. 89: 39-46) on 

 the integration of the gene pool. Here resistance of Drosophila against 

 DDT can be inherited by the combined alleles of 2 loci, either AA/BB or 

 CC./DD. All the other genotypes did not show any resistance, so the 

 interaction of the 2 loci (epistasis) is responsible for resistance in 

 this case. Crosses between the 2 resistant types would yield non- 

 resistant Aa/Bb/Cc/Dd. 



GERHOLD: Dr. Stern, apparently most of your remarks concerning 

 application of quantitative genetic theory pertain to improvement of 

 populations that are in genetic balance. Early in your paper you recom- 

 mended usefulness of species hybrids. Are not the genetic systems of 

 the first few hybrid generations quite disrupted? Would you comment on 

 quantitative genetic theory to handle such situations? 



STERN: I have not covered any quantitative genetic theory here 

 except some applying to resistance. It's quite clear that work with 

 species hybrids has some quite severe limitations on the side of adapta- 

 tion, growth and like things. However, it might be possible to introduce 

 genes from exotic species having an effect on general (horizontal) 



