The Challenges 



A total of 5,543 challenges were made in the course of 

 delineating the population of isolates into clones. Because 

 the blind challenges contained 192 isolate x self chal- 

 lenges, the precision of the methodology could be investi- 

 gated. The challenges were classified into two groups 

 where compatibility (yes) is expected (isolate x self and 

 isolates within clones) and two groups where noncom- 

 patibility (no) is expected (clones in plots and plots x 

 plots). Because they were expected to be identical, "yes" 

 was expected any time an isolate was selfed. "Yes" was 

 also expected when two isolates shown by previous tests 

 to belong to the same clone were challenged (verifying 

 region of the Punnett square). "No" was expected if two 

 clones separated by more than 500 meters were chal- 

 lenged according to current ideas about clone extent in 

 western forests (Anderson and others 1979). "No" was ex- 

 pected if two previously defined clones from the same plot 

 were challenged. Two kinds of disagreement were in- 

 vestigated: the readers recorded independently and could 

 disagree with each other, or they could agree (in the read) 

 on thf> result of the challenge but disagree with the ex- 

 pected result. The selfed isolates provide a control or 

 check on the challenge methodology because the only 

 presumed source of variation is the challenge system. The 

 readers disagreed 4 percent of the time when a "yes" was 

 expected from selfed challenges. The population of reads 

 obtained from the clone-verification zones of the Punnett 

 squares was virtually the same as reading selfed isolates 

 (table 1). 



When "yes" was expected from blind reading of selfed 

 isolates, the readers agreed to disagree with the expecta- 

 tion 5 percent of the time. The clone-verification popula- 

 tion was not significantly different from the selfed isolates 

 (table 2), even though the former population showed 

 reduced disagreement. 



Clones will fail to grow together because they belong to 

 different intersterility groups or they are different geno- 

 types within the same group. Thus the level of reader 

 differences with a "no" expectation was assessed. Dis- 

 agreement among readers was significantly different 

 between clones within plots and clones between plots 

 (table 3), yet the difference was not significant when 

 readers agreed to disagree with expected results (table 4). 



Geographic Extent of Clones 



The geographic extent of clones within a National 

 Forest was examined by constructing challenges of coded 

 isolates, in all possible combinations of isolates, obtained 

 from the 12 National Forests. The challenges between 

 clones from different plots resulted in 95 percent "no," 

 2 percent "yes," and 3 percent "maybe" (table 5). The 

 National Forests varied from percent to 8 percent 

 double "yes" when it was expected that both would be 

 "no." 



Table 1— Disagreement between two readers of challenges 



between field isolates of Armillaria when isolates were 

 expected to be compatible 





Number of challenges 





Isolate 



Reader 



populations 



Yes and no Two yes Total 



disagreement 







Percent 



Selfed 



7 176 183 



3.8 



Clone verification 



17 427 444 



3.8 



= 0.05 Probability of larger (1 df) = 0.90 



Table 2— Accord in dissent with theoretical expectation between 

 two readers of challenges between field isolates of 

 Armillaria when isolates were expected to be compatible 





Number of challenges 





Isolate 





populations 



Two no Two yes Total 



Reader dissent 







Percent 



Selfed 



9 176 185 



4.7 



Clone verification 



16 427 443 



3.6 



= 0.26 Probability of larger x^ (1 df) = 0.70 



Table 3 — Disagreement between two readers of challenges 



between field isolates of Armillaria when isolates were 

 expected to be incompatible (challenges from verify 

 zone-no region of clone-definition Punnett squares and 

 clones between plots) 



Number of challenges 



Isolate Yes Reader 



populations and no Two no Total disagreement 



Percent 



Verify zones 25 426 451 5.5 



Clones between plots 102 3,106 3,308 3.1 



x' = 4.00 Probability of larger x^ (1 df) = 0.05 



Table 4 — Accord in dissent with theoretical expectation between 

 two readers of challenges between field isolates of 

 Armillaria when isolates were expected to be incom- 

 patible (challenges from verify zone-no region of clone- 

 definition Punnett squares and clones between plots) 



Isolate 

 populations 



Number of challenges 

 Two yes Two no Total 



Reader 

 dissent 











Percent 



Verify zones 



17 



426 



443 



3.8 



Clones between plots 



95 



3,128 



3,223 



2.9 



= 0.76 Probability of larger x^ (1 df) = 0.40 



6 



