INOCULATION OF WESTERN WHITE PINE WITH BLISTER RUST 359 



PROGENY TEST DESIGNS 

 EARLY TESTS 



Between 1950 and 1953, at the start of our work on blister rust 

 resistance in western white pine, we pollinated to the limit of female 

 flowering on available candidate trees (i.e., rust-free or phenotypically 

 resistant trees selected in heavily infected natural stands) . This kept 

 us quite busy, but resulted in a very irregular mating scheme among full- 

 sib seed progenies that were scheduled to be sown, inoculated and 

 examined in a given progeny test. Some tests included 20 progenies of 

 one candidate, and only one or two of another candidate. 



The 30 to 100 full-sib seed progenies produced per year in this 

 period were stratified during winter and sown each spring in annual 

 progeny tests between 1952 and 1955. Tests consisted of 9 complete 

 randomized blocks with each progeny represented by one 10-seedling row 

 plot replicate per block. Seeds for a given row-plot were sown in 10 

 2"x2" tar paper plant bands, 6 row plots per flat [Fig. 2), with the flats 

 then plunged, back-to-back to ground level, in nursery beds. The largest 

 of these four tests contained about 9,000 seedlings. These were inocu- 

 lated at 2 years of age, in three successive inoculation runs of 3,000 

 seedlings each. Canvas inoculation chambers were used in the shade of 

 canvas flies (Fig. 1A) . 



It soon became apparent that the mating design, the canvas inocula- 

 tion chambers, and several inoculation runs of the early tests were 

 unsatisfactory. The fragmentary mating design often left us quite 

 puzzled as to which candidates transmitted useful levels of resistance. 



RECENT TESTS 



In tests sown from 1960 on we increased the number of seedlings in 

 each row-plot to 16, and the number of blocks to 10. Also, we used a 

 tester-mating scheme wherein each candidate was crossed with four tester 

 trees. Because of these improvements and the increased efficiency of a 

 larger pollination crew, test size increased markedly. The seed progenies 

 sown each year ranged from 160 to 475 and occupied from 200 to 800 runninj 

 feet of nursery bed space. We no longer had time nor money for handling 

 plants in plant bands and flats, or for stratifying the many different 

 seed lots; instead, in the fall we presowed seed for each progeny on a 

 drop of methyl-cellulose mucilage (with "Captan" fungicide added) placed 

 on stenciled seed planting spots on paper-towel strips (Fig. 3A) . Length 

 of each strip was equivalent to 10 3"x24" row-plots (Fig. 3B) . Each of 

 these 10-row-plot seed strips (replicates) was cut apart and prerandomized 

 for planting the 10 blocks. Each 3"x24"strip was then simply laid on the 

 surface of the bed at the correct row and block position and covered with 

 1/4" of plasterer's sand. Normally, germination occurred the following 

 spring, and tests were inoculated either that same fall when earliest 

 germinating seedlings were 1 year old or the following fall when they 

 were 2 years old. 



