606 V. STEEN ACKERS 



the other white pine hybrids, but the vigor is not too bad. Also, 

 multiple trait selection may be feasible. It would be impossible to 

 combine selection for growth and weevil resistance at the same time once 

 the trees were weeviled, because it is very difficult to assess height 

 growth of weeviled trees . But we can select for different traits at 

 different ages in the same tree, assuming that it's possible to evaluate 

 height growth at an early age. Perhaps the trees could be sprayed for a 

 few years to protect them from the weevil; during this period, one could 

 assess height growth, then stop spraying and test for weevil resistance. 

 This would have to be done on a very uniform site to avoid environmental 

 variations which would reduce the effectiveness of selection for height 

 growth, but it could be done on a small area, especially if caging 

 techniques were used. These are a few ideas I'd like to throw out for 

 consideration. 



HEIMBURGER: I believe that it would be possible to do both except 

 that frequency of blister rust resistance in P. strobus is very low. 

 It's about one in 10,000 to one in 1,000, and therefore, you have to 

 start with very large numbers, and the same thing with weevil resistance. 

 It depends very much on the provenance. Some provenances are much more 

 weevil resistant than others. All the materials from eastern North 

 America near the Atlantic are much more weeviled than the western 

 provenances of P. strobus. There leader thickness is a very important 

 factor which can be very simply measured in the nursery stage, and there 

 is a good correlation between leader thickness of adult trees and leader 

 thickness of seedlings. We have one provenance where we collected seeds 

 from a tree, and we measured the leader thickness of the seedlings and 

 they were all much thinner in this population. So, I would stick to 

 scheme two of Gerhold but would add to it that we need plenty more. That 

 the frequency of resistance is low for blister rust and quite variable in 

 relation to weevil, depending on the provenance. 



BECKER: I am delighted to hear Dr. Gerhold discuss some animal 

 breeding systems called tandem index selection and culling levels. I 

 have been listening to various people discuss breeding systems, and it 

 appeared to me you have been using the cereal crop systems as your model, 

 and yet in working here in Moscow, I have always had the feeling that 

 tree breeding was much closer to animals than it was to the cereal crops. 

 I know this may shock you a bit, but I consider trees out in the Far 

 East as being bull No. 1. Now, I'd like to comment on these three systems 

 that you have proposed. I would say that the tandem method would be 

 almost an impossibility. You have too long a generation time to do 

 this. Tandem method means you would select for growth for say five 

 generations, then you would shift to selecting for blister rust resistance 

 for five generations, then you would shift to weevil resistance for five 

 generations. Well, this seems to be much too long, and actually, in 

 animal breeding, there is very little of this going on. This was just a 

 suggestion thrown out by Hazel way back in the early days. Never really 

 practiced by animal breeders. The next one would be index selection, 

 and I believe this might be possible if you had full-sibs in three 

 different areas where you more or less simultaneously subjected one group 

 to blister rust exposure and another group to weevil exposure, and another 

 group that you measured for growth. Then you put it into a grand index 

 and it comes out in a computer that says select such and such a family. 

 However, this does seem to have some problems too. I don't think you're 

 going to do that, really, from my slight experience with forest tree 

 breeding. So this leaves the other method, which really you have outlined 



