to fruit by PC (i.e., first sampling) was greater in the 

 sides of blocks having 4 trap trees than either 1 or 2 

 trap trees. The same result was found for amount of 

 injury by PC in non-trap trees. For the second sampling, 

 the amount of injury produced by PC to fruit on the 

 trap trees was similar on the different sides of blocks, 

 regardless of the number of trap trees deployed. The 

 amount of injury by PC in non-trap trees was very low 

 in general and was not associated with the number of 

 trap trees deployed on the different sides of a block. 

 Overall, three weeks after deploying 4 BEN and 1 

 GA (i.e., by June 15-16) trap ti"ees were, on average, 

 28 tunes more likely to retlect injury by PC than all 

 perimeter-row non-trap trees sampled in blocks having 

 small trees, and were 15 times more likely to reflect 

 injury by PC than all perimeter-row non-trap trees 

 sampled m blocks having large trees. 



Conclusions 



From the first study, we learned that the effective 

 distance over which trap trees seemed to aggregate PC 

 injury was at least 50-56 yards for the first sampling 

 date and 34-40 yards for the second sampling date. We 

 also determined that trap trees were less able to 

 concentrate injury by PC by the first sampling than by 

 the second sampling, in particular if blocks had large 

 trees. We believe this result is largely due to the cool 

 and rainy weather that prevailed from the moment in 

 which we baited the trap trees (on May 25-26) until 

 the first sampling (on June 1-2). Because during that 

 period of time the release rates of the synthetic lures 

 were presumably low, it is conceivable that PCs may 

 have not been strongly attracted to trap trees and, as a 

 consequence, injury to fruit by PC was more spread 

 fi-om the trap tree for the first sampling date than for 

 the second sampling date. After three weeks, most 

 injury by PC occurred on the trap trees and on the most 



adjacent trees. 



Our results from the second study show that, except 

 for blocks having large trees in the first sampling date, 

 the amount of injuiy by PC to trap trees was similar on 

 perimeter rows having 1 , 2, or 4 trap trees, regardless 

 of the size of trees in a block. This finding indicates 

 that a single trap tree deployed mid-way of a perimeter 

 row of about 120 yards will be sufficient to monitor 

 accurately the seasonal course of injury by PC to fruit. 

 Remarkably, the amount of injury by PC to perimeter- 

 row ixuit located in the side of blocks having no ti-ap 

 trees was as low as the amount of injury that occun-ed 

 in all perimeter-row fruit when we excluded injuiy on 

 the trap trees. Thus, we should not expect greater-than- 

 normal amounts of injury by PC to perimeter-row fruit 

 by having odor-baited trap trees (regardless of the 

 number). 



Whether the attractiveness of a trap tree could be 

 enhanced by adding other types of stimuli so as to 

 increase its ability to congregate PCs is a question that 

 remains to be investigated. Finding a way of enhancing 

 the attractiveness of trap trees could be very valuable, 

 for instance, in the context of potential direct control 

 of PC by confining insecticide sprays to trap trees only 

 (after a whole-block spray). In the context of 

 monitoring of PC injury, an enhanced trap tree might 

 decrease the extent to which PCs penetrate into orchard 

 blocks (especially in blocks having small trees) by 

 holding PCs on perimeter-row trees, thereby making 

 perimeter-row sprays more efficient. 



A ckn o wiedgm en ts 



We thank Bruce Carlson, Aaron Clark, Dave 

 Chandler, Don Green, Andy Martin, Mo Tougas, and 

 Steve Ware, for allowing us to work on their orchards. 

 Funds for this study were provided by the New England 

 Tree Fruit Research Committee. 



%i« «1> «1* «1« *1^ 

 •^ #^ •y* •^ *y» 



Fruit Notes, Volume 69, Fall, 2004 



17 



