INOCULATION OF WESTERN WHITE PINE WITH BLISTER RUST 367 



The likely source of our problem is apparent * from a look at variation 

 in infection intensity across a large test, as diagrammed in Figure 7 

 1 year after inoculation. There seems to be an obvious association of 

 position of seedlings with degree of infection, especially along the ex- 

 treme east and west walls, and along the south walls of the three con- 

 tinguous inoculation tents. Apparently overheating and drying are still 

 problems to be reckoned with. 



These fluctuations are creating relatively high variances between the 

 blocks and they explain why we use both a relatively large number of 

 replicates (10) and seedlings (160) for each test progeny. Effects of 

 this variation are discussed elsewhere in these proceedings by Becker and 

 Mars den. 



Levels of rust infection diagrammed in Figure 7 held 1 year after 

 inoculation; they were based mainly on needle lesions. Drs . McDonald 

 and Hoff of this Laboratory point out that because of the length of time 

 (50 days, September 1-October 20) required to move across the main test 

 (below the dotted line of Figure 7); moving from east to west, the lower 

 infection on the west may be partly of an artificial nature. They sug- 

 gest this because when they sampled blocks 4-7 in mid-June they found 

 foliage lesions on 99.6% of the sample seedlings. Yet when we reexamined 

 the seedlings of the same four blocks 3 months later in mid-September 

 the infection was much lower (see Blocks 4-7, Figure 7). McDonald also 

 found that premature shedding of infected needles is a resistance trait 

 that is under monogenic control (McDonald, 1969). We know that during 

 warm seasons some infected foliage begins dying and becomes unrecog- 

 nizable as such by early July. 



PLANT SIZE IN RELATION TO SUSCEPTIBILITY 



Starting with our earliest inoculations we noticed the relatively 

 low intensity of needle spotting in certain small, slow-growing, and 

 apparently low-vigor seedlings and progenies. Some of these runty seed- 

 lings were transplants, replacing missing plants; others were located in 

 under-fertilized portions of seed beds, and still others were from late 

 germinating seed or in self-pollinated progenies. In one progeny test, 

 simple correlation of number of needle infections on a 1,715 lineal cm 

 needle sample with average progeny height (2-year-old seedlings) showed 

 that needle lesion frequency was significnalty , if not strongly, corre- 

 lated with progeny height (r = .403, significant at the 5% level). 

 Regression coefficient, "b", was roughly 430 needle spots per 1,715 lineal 

 I cm needle sample, for each additional cm of plant height. 



And, as expected in the same test progeny (Squillace and Bingham, 

 1958) seed weight was also significantly, if not strongly associated with 



4< average progeny 2-year-old seedling height (r = .426, significant at the 

 5% level, b = .00025 cm of height for each mg of seed weight). Would 

 it be necessary to adjust observed needle-lesion-intensity values for the 

 extraneous height variation associated with variation in seed weight? For 



. practical purposes, the adjustment proved to be unnecessary. Average 



progeny seed weights seldom ranged 100 mg from the mean (230 mg ) . Even 

 t if they were that far from the average, the increment of seedling height 

 due to seed weight would have been relatively small (i.e., 100 x .00025 cm 



jg = 0.025 cm height); and this small increment of height would produce a 

 small change in number of spots per 1,715 lineal cm needle sample (i.e., 

 0.025 x 430 spots = 11 spots, while the average progeny had about 200 

 spots per 1,715 lineal cm needle sample). 



