BREEDING RUST RESISTANT TREES: MODERATOR'S SUMMARY 631 



in a breeding program to calculate heritability and genetic gain, it will 

 be of no value unless wise, adequate sampling is done to select superior 

 parent trees in the main provenances of the species that is to be improved, 



Once a proper choice of parental trees has been made, an aggressive 

 crossing program must be launched to develop a widely diversified gene 

 pool. Large populations of seedlings from this diverse gene pool must 

 be grown and screened economically if the program is to be successful. 

 The nursery screening tests for rust resistance, which we observed 

 yesterday at the Intermountain Forest and Range Experiment Station nursery 

 are an excellent example of an effective first screen. I caution, how- 

 ever, against employing nursery test methods that are designed to identify 

 and save only seedlings with hypersensitivity. Seedlings that survive 

 the nursery screen test should be transplanted to a regional performance 

 nursery located in a high infection area (in an ecological environment 

 conducive to heavy infection) , where rust infection can be intensified 

 by the extensive interplanting of many species of Rites. Aeciospores 

 from many different parts of the region should be used to infect the 

 Ribes. The regional performance nursery approach proposed by Pauley and 

 Ahlgren (these proceedings) would certainly be a good second screen. 



The outstanding saplings which emerge from these two types of tests, 

 and which are candidates for use in seed orchards, should be evaluated, 

 whenever possible, for disease and insect resistance, winter hardiness, 

 growth rate, tree form, and general adaptation in regional, national, 

 and international tests. Species such as pines that can be propagated 

 vegetatively through grafting or cuttings lend themselves to simultaneous 

 testing at many sites much better than do crop plants. The use of a 

 widespread network of national and international testing sites will not 

 only assist in selecting individual trees with unusual disease and 

 insect resistance, but can also lead to the identification of trees with 

 an unusually broad range of adaptation. 



.APPLICATION OF THE ABO\T£ PRINCIP^S IN CEREAL CROPS 



Before closing I would like to show you how some of these principles 

 have been employed in the Mexican - CIMMYT 6 wheat breeding program during 

 the past 10 years to revolutionize wheat production in a number of different 

 countries. The wide adaptation of the Mexican varieties is not accidental, 

 but the result of methods used in their development. The results have 

 destroyed many of the former plant breeding concepts that placed much 

 emphasis on the importance of tightly fixed variety-environment inter- 

 actions which precluded the development of widely adapted varieties. 

 The current results clearly indicate the advantages of broadly adapted 

 varieties with built-in stability for grain yield over a wide range of 

 conditions (sites, elevations, latitudes, and season). I predict that, 

 if a sufficiently diversified gene pool is developed and if an adequate 

 number of testing sites are employed, varieties or cultivars of maize 

 and pines with amazingly broad adaptation can be developed similarly. 



Centro Internacional de Mejoramiento de Maize y Trigo (International 

 Maize and Wheat Improvement Center) . 



