Table 2 — Community status 3 years postspray on the pilot plots 



Average percent canopy cover 



Diversity measurements 



Treatment 



CENMAC 



Other 

 forbs 



Grasses 



Average No. 

 species 



Total No. 

 species 



S-W index 

 average (1 SD) 



Check plots 

 Tordon 



Check plots 

 Tordon 



Tordon+Stinger 

 Stinger 



27 

 3 



30 

 3 

 4 



16 



7 

 6 



20 

 17 

 16 

 22 



Fort Missoula 

 8 

 28 



Lolo 



27 

 46 

 47 

 37 



16.0 

 16.0 



35.3 

 34.3 

 37.7 

 37.0 



22 

 22 



57 

 58 

 59 

 58 



0.76(0.035) 

 .94 (.148) 



1.13 (.118) 



1.26 (.167) 

 1.29 (.060) 



1.27 (.050) 



'CENMAC = Centaurea maculosa Lam. 



as relative standard deviations for paired original 

 and blind synthetic transects. Soil samples submit- 

 ted to the residue analysis laboratory included blind 

 split field duplicates and blind herbicide-free check 

 plot soils. The laboratory used internal QA/QC pro- 

 cedures approved by the Environmental Protection 

 Agency. Those methods included analyses of dupli- 

 cate sample splits, estimating recovery of known ad- 

 ditions to herbicide-free check plot soils from the 

 study sites, and incorporation of reagent blanks in 

 the analytic stream. 



RESULTS AND DISCUSSION 



Both the pilot diversity study and the formal com- 

 munity response study suggested that plant diver- 

 sity could be maintained or even enhanced when 

 these herbicide treatments were used to control 

 spotted knapweed. 



Pilot Diversity Study 



The two pilot diversity study sites, which had been 

 sprayed in 1985, and the community status meas- 

 ured 3 years later indicated that diversity on the 

 herbicide-treated plots was as high as or higher than 

 on the check plots (table 2). Spotted knapweed con- 

 trol of 87 to 90 percent was still being maintained on 

 the Tordon-treated plots but had declined to 53 per- 

 cent on the plots that had been sprayed with Stinger 

 alone. 



Formal Community Response Study 



All six herbicide treatments provided good control 

 of spotted knapweed 1 and 2 years postspray. Little 

 difference occurred between early and late herbicide 

 treatments. The spotted knapweed response to her- 

 bicide type is summarized in table 3 by pooling early 

 and late application dates for all four sites. Absolute 



Table 3 — Mean spotted knapweed canopy cover (absolute 

 percentage) during the warm season and average 

 control as a percentage of untreated check plots 





Prespray 



1 year postspray 



2 years postspray 





(1989) 



(1990) 



(1991) 







Percent 



Percent 





Cover 



Cover reduction 



Cover reduction 



Check 



10.7 



10.1 



10.1 



Tordon 



9.4 



.8 92 



1.1 89 



Stinger 



10.6 



.6 94 



1.0 90 



Curtail 



9.6 



.9 91 



1.6 84 



canopy cover of spotted knapweed in individual 

 treatment plots 1 and 2 years after spraying ranged 

 from less than 0.5 percent to a maximum of 9.2 per- 

 cent, with all but three of the 72 herbicide-treated 

 plots having less than 3 percent spotted knapweed 

 cover. 



We believe that posttreatment year efficacy rates 

 of 99 percent are not realistic for most weed control 

 projects on natural vegetation sites. Weeds will 

 occur where there are skips in the spray swath, 

 droplet-shielded microsites, and resistant individu- 

 als. If land managers require more than 99 percent 

 efficacy, they must conduct a foUowup spot spraying. 



Blind QA/QC rereadings indicated excellent preci- 

 sion in determining the community diversity vari- 

 ables. The average relative standard deviation was 

 5.8 percent for species richness and 5.6 percent for 

 the Shannon- Weaver index (n = 22 rereadings). 



The variance among plots receiving the same 

 treatment within a site was usually small, although 

 some outliers occur in the data base (table 4). 



The check plot sets generally had the highest 

 average prespray diversity values (tables 5, 6). A 

 graphic inspection of the data for the period 1988 

 through 1991 shows that, overall, the impact of the 



3 



