Influence of Host Tree Proximity on Adult Plum Curculio

BEHAVIOR Inﬂuence of Host Tree Proximity on Adult Plum Curculio (Coleoptera: Curculionidae) Responses to Monitoring Traps

TRACY C. LESKEY AND STARKER E. WRIGHT

USDAÐARS, Appalachian Fruit Research Station, 2217 Wiltshire Road, Kearneysville, WV 25430

Environ. Entomol. 33(2): 389Ð396 (2004) ABSTRACT We evaluatedresponses of plum curculio, Conotrachelus nenuphar (Herbst), to baited andunbaitedblack pyramidandPlexiglas panel traps in markÐreleaseÐrecapture experiments to determine whether presence of host apple trees diminishes trap effectiveness. Identical experiments were conductedinan open Þeld(little competition present) andin an unsprayedapple orchard (presence of competition from natural host apple trees). Results of markÐreleaseÐrecapture experiments were successful; we recaptured21.0 and2.9% in the Þeldandorchard,respectively. Plum curculio response to traps was inßuencedby presence of apple trees, because signiÞcantly more recaptures were recordedinthe Þeldcomparedwith the orchard.Pyramidtraps recapturedmore plum curculio in the Þeldcomparedwith the orchardandrecapturedsigniÞcantly more plum curculio comparedwith panel traps. Female plum curculio respondedinsigniÞcantly greater numbers to pyramidtraps baitedwith benzaldehydecomparedwith unbaitedtraps in the Þeld,but not under orchard deployment. Combined data indicate that positioning traps in proximity to host apple trees diminishes both trap and bait effectiveness either by direct competition from host tree stimuli or by physical obstruction of trap-related visual stimuli and/or volatile release. We conclude that an effective trap-basedmonitoring system for plum curculio will require better odorbaits, trapping mechanisms, and/or deployment strategies that can overcome visual and olfactory interference presentedby host apple trees.

KEY WORDS plum curculio, Conotrachelus nenuphar, monitoring traps, grandisoic acid, benzaldehyde

IDENTIFICATION OF ATTRACTIVE HOST plant volatiles and eastern UnitedStates (Yonce et al. 1995). Lack of a pheromones usedby phytophagous pest species has reliable monitoring strategy to detect plum curculio become an important component of behaviorally appearance and/or abundance in orchards to deter- basedmonitoring strategies (Foster andHarris 1997). mine needandtiming of insecticidesprays (Prokopy In combination, host plant volatiles often synergize or andCroft 1994) has ledto trap developmentand enhance insect responses to sex and/or aggregation deployment for plum curculio (Tedders and Wood pheromones (Landolt and Phillips 1997). These en- 1994; Yonce et al. 1995; Mulder et al. 1997; Prokopy and hancedresponses have been recordedfora number of Wright 1998; Prokopy et al. 1999, 2000; Johnson et al. species belonging to the family Curculionidae 2002; Leskey andProkopy 2002). Furthermore, iden- (Landolt1997) andhave ledto successful monitoring tiÞcation of potential baits for traps basedon attractive systems for several species, including Rhynchophorus host plant volatiles (Leskey andProkopy 2000, Leskey palmarum (L.) (Oehlschlager et al. 1993), Rhyn- et al. 2001, Prokopy et al. 2001) in conjunction with the chophorus cruentatus (F.) (Giblin-Davis et al. 1994), attractive component of the male-produced aggrega- and Metamasius hemipterus sericeus (Olivier) (Giblin- tion pheromone, grandisoicacid(Eller andBartelt Davis et al. 1996). 1996) also continues to be an area of active research. The plum curculio, Conotrachelus nenuphar Several trap types have been testedfor plum cur- (Herbst), is a major pest of stone andpome fruit in culio. The pyramidtrap is believedto providean eastern andcentral North America (Racette et al. attractive visual stimulus by mimicking a tree trunk 1992, Vincent et al. 1999) andis one of the most (Tedders and Wood 1994, Mulder et al. 1997) and has destructive pests of peaches and plums in the south- been reportedto capture more crawling than ßying individuals (Prokopy and Wright 1998). Panel traps do not have a speciÞc visual cue associatedwith them and Mention of trade names or commercial products in this publication is solely for the purpose of providing speciÞc information and does not are designed to capture ßying adult plum curculio imply recommendation or endorsement by the U.S. Department of (Prokopy et al. 2000). The border row of fruit orchards Agriculture. would seem to be the ideal location for deployment of 390 ENVIRONMENTAL ENTOMOLOGY Vol. 33, no. 2 these monitoring traps because this is where adults pheromone (grandisoic acid), benzaldehyde in com- Þrst enter the orchardandoviposition damageÞrst bination with grandisoic acid, or unbaited controls occurs (Racette et al. 1992). However, in studies in were recorded at both study sites to assess the impact which pyramid and/or panel traps were deployed in of host apple trees on recapture of plum curculio in the border row of untreated apple orchards, trap cap- traps. To be effective, a trap-basedmonitoring system tures did not reßect amount of oviposition injury ob- must be able to overcome the natural inference pro- servedin fruit trees (Prokopy et al. 1999, 2000; Leskey vided by host apple trees. If presence of apple trees has andWright 2004), andhence failedto serve as a no effect on monitoring traps, we predict that trap reliable tool to determineneedfor andtiming of in- captures will be statistically equal between orchard secticide application. A stronger correlation between andÞeldlocations. trap captures andoviposition injury has been recorded in peach orchards in Arkansas and Oklahoma Materials and Methods (Johnson et al. 2002). A potential explanation for the poor relationship Study Sites. Studies were carried out within an un- between trap captures andoviposition injury ob- sprayedsection of a commercial apple orchardandan servedin these previous studiesis that presence of open Þeld to provide study sites with and without host apple trees reduces the likelihood that foraging competition from and/or physical disruption by host plum curculio will locate andenter monitoring traps apple trees. The 6-ha commercial orchardwas located due to physical, visual, and/or olfactory inßuence of in Inwood, WV, and planted in 1988. Trees were me- host trees on baitedmonitoring traps. For example, dium-size ÔEmpireÕ on M.9 rootstock planted Ϸ3m Prokopy et al. (2000) reportedthat clear Plexiglas apart within rows and5 m between rows. The 2.5-ha panels baitedwith synthetic fruit volatiles (ethyl Þeld,plantedin mixedfescue, was locatedat the Ap- isovalerate andlimonene) placed2 m from the woods palachian Fruit Research Station in Kearneysville, andsome distancefrom the orchardwere signiÞcantly WV. A woodlot borderedtheÞeldto the north, more attractive than unbaitedpanels, but they found whereas hedgerows bordered the Þeld to the west and no differencebetween baitedandunbaitedpyramid south, anda fence separating pastoral landwith sev- traps placednext to apple tree trunks or baitedand eral rows of unmanagedapple trees borderedtheÞeld unbaitedcylindertraps placedin apple tree canopies to the east. Groundcover at both sites was trimmed to in the border row. Furthermore, Pin˜ ero et al. (2001), a height of 10 cm for these studies to decrease the reportedthat clear sticky Plexiglas panels andblack potential inßuence of other nonhost vegetation. Sites pyramidtraps placedin the borderregion between the were locatedwithin 15 km of each other andexperi- woodsandan unsprayedorchardcaptureda signiÞ- encedsimilar weather conditions. cantly greater number of plum curculio when baited Trap Types. We evaluatedtwo trap types in both with a synthetic fruit volatile, benzaldehyde, in com- locations: 1) black pyramidtraps originally designed bination with grandisoicacidcomparedwith any to monitor pecan weevil, Curculio caryae (Horn), pop- other bait combination or to unbaitedtraps (with the ulations (Tedders and Wood 1994) but modiÞed to exception of ethyl isovalerate andgrandisoicaciddur- monitor plum curculio (Prokopy andWright 1998, ing the early stages of plum curculio immigration). Prokopy et al. 1999); and2) Plexiglas panel traps However, how plum curculio responded to these same (Dixon et al. 1999, Prokopy et al. 2000, Pin˜ ero et al. bait combinations in the border row of an orchard was 2001) coatedon one sidewith Tangletrap (GemplerÕs, not tested. Thus, plum curculio responded in signiÞ- Belleville, WI). We deployed two trap types to allow cant numbers to baitedpanel traps (Prokopy et al. capture of both crawling (pyramidtraps) or ßying 2000) andbaitedpanel andpyramid(Pin ˜ ero et al. (panel traps) plum curculio, although pyramidtraps 2001) traps comparedwith unbaitedtraps when traps also may capture ßying adult plum curculio (Prokopy were deployed adjacent to apple orchards, but not to andWright 1998, Dixon et al. 1999). baited traps in the border row of apple orchards them- Baits. We evaluatedtraps baitedwith one of four selves (Prokopy et al. 2000, Leskey andWright 2004), bait treatments: 1) the synthetic fruit volatile benzal- indicating that presence of host apple trees may re- dehyde (Aldrich, Milwaukee, WI), found to be at- duce trap effectiveness in or near the orchard. tractive to plum curculio (Pin˜ ero et al. 2001, Prokopy The aim of this study was to determine whether et al. 2001); 2) grandisoic acid (ChemTica Interna- presence of host apple trees is responsible for the tional, S.A., San Jose, Costa Rica), a synthetic version inability of baitedpyramidandpanel traps to capture of an attractive component of the male-produced plum curculio after fruit set andprovidean explana- pheromone (Eller and Bartelt 1996); 3) benzaldehyde tion for the inability of trap captures to reßect accu- in combination with grandisoic acid; and 4) an un- rately the amount of oviposition injury observedin baitedcontrol treatment. Release rate of benzalde- orchards (Prokopy et al. 2000). MarkÐreleaseÐrecap- hyde dispensers (500 ␮l of benzaldehyde in 1-ml ture experiments using plum curculio collectedfrom white, UV-resistant, low-density polyethylene vials wildpopulations were conductedattwo sites: within (Wheaton ScientiÞc Products, Millville, NJ) was an unsprayedapple orchardandwithin an open Þeld. Ϸ10 mg/d; release rate was determined by exposing 25 Responses to Plexiglas panel andblack pyramidtraps dispensers to open-air outdoor conditions at 25Ð30ЊC baitedwith either a synthetic fruit volatile (benzal- for 96 h and weighing each dispenser daily to deter- dehyde), a synthetic male-produced aggregation mine milligrams of compoundlost per dayover the 4-d April 2004 LESKEY AND WRIGHT:INFLUENCE OF HOST TREE 391

Fig. 1. Experimental layout of alternating pyramidandpanel traps at the borderof circular subplots of increasing diameter (5, 10, and20 m) in (A) Þeldand(B) apple orchard(stripes represent tree rows within the unsprayedapple orchard). period. Release rate of grandisoic acid dispensers Plum Curculio Adults. Plum curculio usedin these (25 mg) was determined by the manufacturer to be experiments were collectedin Jefferson andBerkeley Ϸ1 mg/d. For pyramid traps, baits (a single benzal- counties in West Virginia. In late May, these overwin- dehyde dispenser and/or a single grandisoic acid lure) teredadultswere collectedfrom populations present were placedwithin the boll weevil trap collection in unsprayedfruit trees using beating sticks andcol- device located at the top of each trap. For panel traps, lection sheets placedbeneath tree canopies. Plum a single benzaldehyde dispenser was attached to the curculio were assumedto be approximately the same edge of each panel using a locking plastic cable tie age, segregatedaccordingto sex within 24 h after and/or a single grandisoic lure was attached to the collection by using criteria described by Thomson upper right-handcorner of the panel with a small (1932), andheldas separate sexes in wax-coatedcups binder clip. (473 ml) with clear plastic lids at 25ЊC under a pho- Block Setup. We established55 by 20-m blocks in toperiod of 16:8 (L:D) h to mimic long-day conditions both the open Þeldandapple orchard.Within each in nature from late May to late June. On average, 40 block, three sets of alternating pyramidandpanel single-sex individuals were housed within a cup. Fresh traps (four of each trap type in each set) were installed apple fruit was providedalongwith a wettedcotton in three circular subplots of increasing diameter (5, 10, and20 m) (Fig. 1). Traps were spacedequidistantly wick as a water source. Although some herbivorous along the perimeter of each circular subplot. Each of insects respondto olfactory stimuli differentlybased the three circular subplots was divided into four quad- on diet provided before experimentation, no such re- rants; each quadrantcontaineda pyramidandpanel sponse was observedwith plum curculio in previous trap baitedwith one of the four baits being tested. studies involving plum curculio attraction to synthetic Within the apple orchardsubplots, trees that hadtraps host volatile attractants (Leskey andProkopy 2001). locatednext to them within the tree row were pruned Before release, adults were starved for 24 h [because to accommodate the trap and provide Ϸ1 m of open this protocol was previously shown to enhance re- space. Within the Þeldblock, the potential inßuence sponsiveness to olfactory stimuli with no negative from host trees was minimizedby providingat least a physiological effects (Prokopy et al. 1995)] and 100-m buffer surrounding the test block that was free markedwith a small uniquely coloreddotof paint on of any known host tree species of plum curculio their elytra to designate sex, release point, and sample (Maier 1990). interval. This marking technique was similar to that 392 ENVIRONMENTAL ENTOMOLOGY Vol. 33, no. 2

traps were rotatedclockwise by one quadrantto ac- count for any positional bias andbaits were replaced. All releases were made after petal fall (1 May) because it is after this phenological event when plum curculio captures by baitedtraps begin to declineanddevel- oping fruit are at risk for oviposition injury (Prokopy et al. 2000). Releases were made on 16 May, 23 May, 27 May, and4 June 2001 when apple fruit were Ϸ12, 20, 22, and30 mm in diameter,respectively, at Ϸ1800 hours at both locations. Data Analysis. Analysis of covariance (ANCOVA) was conducted using the SAS PROC GLM procedure (SAS Institute 2001) (treating subplot size as a covariate) to determine which class variables (location, trap type, bait, sex, release event, andquadrant)sig- niÞcantly inßuencedplum curculio recaptures across andwithin orchardandÞeldsites. This particular analysis provided a means to test the effects of all qualitative class variables on the dependent variable (plum curculio recaptures) while controlling for ef- Fig. 2. Release device used to facilitate plum curculio fects of variation contributedby the quantitative vari- movement from central release point in each circular subplot. Release device consisted of (A) wooden dowels in- able (subplot size) (SAS Institute 2001). For those sertedinto a (C) ßorist foam plug heldwithin a (D) plastic factors foundto be signiÞcant, datawere analyzedfor cup. Chilledplum curculio were releasedin conjunction with that particular factor using a two sample t-test (loca- a (B) paper towel to provide shelter and diminish escape tion, trap type, andsex) to determinewhether signif- responses. icant differences existed between means. The relationship between temperature (daily maximum, minimum, andaverage) anddailyrecaptures within usedby Butkewich andProkopy (1997) in markÐre- the Þeldandthe orchardwas analyzedin separate leaseÐrecapture experiments of adult plum curculio. analyses (treating temperature as a quantitative vari- Before release, plum curculio were chilledfor able using the SAS PROC GLM procedure); none of 30 min inside wax-coated paper cups with moistened these models were signiÞcant. Analysis of variance paper towels over an ice bucket to diminish escape (ANOVA) andTukeyÕs honestly signiÞcant difference responses. Plum curculio were releasedin groups of 40 (HSD) were conducted using the SAS PROC GLM (20 males and20 females) in the center of each of the procedure (SAS Institute 2001) to determine whether three circular subplots. The contents of each cup either male or female plum curculio responded sig- (plum curculio andpaper towels) were gently re- niÞcantly to any particular bait deployed with either movedandplacedin the center of a release device pyramidor panel traps within orchardandÞeldsites consisting of a modiÞed plastic cup with attached at each distance. wooden dowels inserted into a plug of ßorist foam (Fig. 2). Wooden dowels promoted plum curculio Results movement from the release device by providing a substrate for either crawling or ßight. With the ex- Recapture Proﬁles. The percentages of marked ception of two plum curculios in the open Þeldblock, plum curculio recapturedwere 21.0 and2.9% in the all plum curculios movedfrom the release devicesinto Þeldandin the orchard,respectively, across all re- subplots. leases. Within the Þeld, recapture rates were 39.4, 15.6, All recoveredfemales were dissectedinthe labo- and8.0% for the 5-, 10-, and20-m subplots, respec- ratory to determine mating status and sexual maturity. tively. Within the orchard, recapture rates were 7.5, Female mating status was determined by removal of 1.3, and0.0% for the 5-, 10-, and20-m subplots respec- the spermatheca for immediate wet mount on a glass tively. More than 75% of all recapturedadultswere slide to establish presence or absence of sperm using recoveredwithin 2 dof release in all subplots located a Nikon SMZ 1500 (7.5Ð112.5ϫ) stereomicroscope. in both the Þeldandin the orchardwith one excep- Sexual maturity was determined by criteria described tion:10-m subplot locatedwithin the orchard(Table by Smith andSalkeld(1964), speciÞcally presence of 1). Within the orchardblock, 11 wildadultswere mature eggs in calyx anddevelopingoocytes in the capturedindicatingthat the mean Ϯ SE size of the vitellarium. wildpopulation within that area was 263.50 Ϯ 43.96 Trap Inspection. Traps were inspected24, 48, 72, according to the Lincoln Index. and96 h after release at Ϸ1800 hours. Recaptured Across Sites. We recaptureda total of 115 plum plum curculio were removedfrom traps andtaken curculio. The general linear model for trap recaptures back to the laboratory to determine mating status and was signiÞcant (F ϭ 9.50; df ϭ 13, 370; P Ͻ 0.001). sexual maturity of all females recovered. Before the Release event, quadrant,andsex of releasedplum next release of plum curculio in each circular subplot, curculio did not have signiÞcant effects on recapture April 2004 LESKEY AND WRIGHT:INFLUENCE OF HOST TREE 393

Table 1. Cumulative percentage of mean recovery and cumulative number of marked plum curculio adults recaptured in ﬁeld and orchard subplots over 4-d recovery period

FieldOrchard 5m%(n) 10m%(n) 20m%(n) 5m%(n) 10m%(n) 20m%(n) Day 1 60.3 (38) 48.0 (12) 46.2 (6) 58.3 (7) 0.0 (0) 0.0 (0) Day 2 81.0 (51) 88.0 (22) 76.9 (10) 91.7 (11) 50.0 (1) 0.0 (0) Day 3 95.2 (60) 100.0 (25) 76.9 (10) 91.7 (11) 100.0 (2) 0.0 (0) Day 4 100.0 (63) 100.0 (25) 100.0 (13) 100.0 (12) 100.0 (2) 0.0 (0)

The n value at day 4 for each subplot reßects total number recaptured out of 160 released individuals. results (Table 2). Location, however, was signiÞcant Within Field. We recaptureda total of 101 plum (Table 2) with more plum curculio recapturedper curculio at our Þeldsite. The general linear modelfor trap in the Þeld(mean Ϯ SE, 0.53 Ϯ 0.08) than in the trap recaptures was signiÞcant (F ϭ 6.35; df ϭ 12, 179; orchard(0.07 Ϯ 0.02) (t ϭ 5.787; df ϭ 1, 382; P Ͻ P Ͻ 0.0001). Sex, release event, quadrant, and bait did 0.0001). Trap type also was signiÞcant (Table 2) with not have signiÞcant effects on trap captures across all signiÞcantly more plum curculio recapturedin pyra- subplots (Table 4). However, trap type was signiÞcant midthan in panel traps (0.54 Ϯ 0.08 and0.06 Ϯ 0.02, (Table 4) with more plum curculio recapturedin respectively) (t ϭ 7.06; df ϭ 1, 382; P Ͻ 0.0001). Bait pyramid(mean Ϯ SE, 0.96 Ϯ 0.13) than in panel traps type was not signiÞcant across sites (Table 2) andthe (0.09 Ϯ 0.03) (t ϭ 6.296; df ϭ 1, 190; P Ͻ 0.0001). Again, covariate, subplot size, was signiÞcant (Table 2) with the covariate, subplot size, was signiÞcant (Table 4); more plum curculio recaptures in the 5-m-diameter more plum curculio were recapturedper trap in the subplots (0.58 Ϯ 0.01) comparedwith 10- or 20-m 5-m-diameter subplot (0.98 Ϯ 0.18) comparedwith 10- subplots (0.21 Ϯ 0.05 and0.01 Ϯ 0.03, respectively). and20-m subplots (0.39 Ϯ 0.10 and0.20 Ϯ 0.06, re- Within Orchard. Within the orchard, we recap- spectively). More than 58% of all recaptures in the tureda total of 14 plum curculio. The general linear Þeldwere in pyramidtraps positionedin the 5-m model of trap recaptures was signiÞcant (F ϭ 2.65; df ϭ subplot. Within the same subplot, recaptures across 12, 179; P ϭ 0.0027). Bait, release event, andquadrant pyramidtraps containing differentbaits were signiÞ- did not have signiÞcant effects on trap recaptures cantly different for females, but not for males (Table (Table 3). However, the effect of sex was signiÞcant 5). SigniÞcantly more females were recapturedin pyr- (Table 3) with more males capturedper trap (mean amidtraps baitedwith benzaldehydecomparedwith ϩ SE, 0.11 Ϯ 0.04) than females (0.03 Ϯ 0.02) (t ϭ unbaitedtraps (Table 5). At all other distances,re- 1.951; df ϭ 1, 190; P ϭ 0.0526). Trap type also was captures from baitedandunbaitedpyramidandpanel signiÞcant (Table 3) because more plum curculio traps for either females and/or males were not signif- were recapturedin pyramid(0.11 Ϯ 0.04) than in icantly different. panel traps (0.03 Ϯ 0.02) (t ϭ 1.951; df ϭ 1, 190; P ϭ Mating Status. We recaptureda total of 57 females, 0.0526). Finally, there was a signiÞcant effect of the 54 in the Þeldandthree in the orchard.We were able covariate, subplot size (Table 3) with more recaptures to determine mating status and sexual maturity of 51 in the 5-m subplot (0.19 Ϯ 0.06) than in the 10 m females recoveredfrom the Þeldandthree females (0.03 Ϯ 0.02) and20 m (0.00 Ϯ 0.00) subplots. Re- recoveredin the orchard;of these, 92 and100% were captures from baitedandunbaitedpyramidor panel matedandsexually mature, respectively. traps for either females and/or males were not signif- icantly different at any distance. Discussion We captureda total of 11 wildindividualsin the orchard,10 of which were capturedin baitedtraps The objective of our study was to determine (eight in pyramidandtwo on panel traps). However, whether presence of host apple trees compromises the no signiÞcant difference was detected among baits. attractiveness andeffectiveness of baitedandun- baitedtraps for plum curculio, resulting in perilous Table 2. One-way ANCOVA for effects of class variables on recapture results of marked-released adult plum curculio across Table 3. One-way ANCOVA for effects of class variables on ﬁeld and orchard sites recapture results of marked-released adult plum curculio in orchard subplots Test Variable F value df P valuea Test variable F value df P valuea Bait 1.64 1,370 0.1786 Location 39.77 1,370 Ͻ0.0001 Bait 1.12 3,179 0.3393 Quadrant 0.68 3,370 0.5661 Quadrant 0.26 3,179 0.8540 Release event 0.97 3,370 0.4059 Release event 1.81 3,179 0.1449 Sex 0.01 1,370 0.9423 Sex 4.14 1,179 0.0428 Subplot sizeb 30.30 1,370 Ͻ0.0001 Subplot sizeb 13.97 1,179 0.0002 Trap type 43.51 1,370 Ͻ0.0001 Trap type 4.14 1,179 0.0428

a Probability of a type III error. a Probability of a type III error. b Covariate. b Covariate. 394 ENVIRONMENTAL ENTOMOLOGY Vol. 33, no. 2

Table 4. One-way ANCOVA for effects of class variables on tical between Þeldandorchard(trimmedto a height recapture results of marked-released adult plum curculio in ﬁeld of 10 cm), the host trees may be interfering with the subplots visual stimulus provided by traps. Perhaps natural vi- Test variable F value df P valuea sual stimuli of host apple trees such as spectral qual- ities and/or speciÞc dimensional characteristics Bait 1.44 3,179 0.2330 Quadrant 0.64 3,179 0.5919 (Prokopy andOwens 1983) were more attractive to Release event 0.48 3,179 0.6939 plum curculio than stimuli provided by black pyramid Sex 0.52 1,179 0.4762 traps. Alternatively, the presence of dense apple fo- Subplot sizeb 23.98 1,179 Ͻ0.0001 Ͻ liage, limbs, andtrunks in the orchardsimply may have Trap type 44.35 1,179 0.0001 obscuredthe pyramidtrap, from the line of vision of a Probability of a type III error. plum curculio. Under either circumstance, the results b Covariate. highlight a decrease in effectiveness of this particular visual cue in the context of an orchard. Host apple tree odor is known to be attractive to reduction of captures under orchard conditions. In- plum curculio (Butkewich andProkopy 1997, Leskey deed, we recorded signiÞcantly fewer recaptures in and Prokopy 2000). Benzaldehyde, the synthetic at- the orchardcomparedwith the Þeld(Table 2) with Ϯ tractant used in our study, was identiÞed from both nearly an eightfolddifferencebetween the mean SE apple blossoms (Buchbauer et al. 1993) andimmature number of plum curculio capturedper trap in the Þeld Ϯ Ϯ plum (Leskey et al. 2001) andwas foundto be attrac- (0.53 0.08) comparedwith the orchard(0.07 tive to plum curculio under Þeld conditions, especially 0.02). Our results indicate that plum curculio are less when combinedwith the aggregation pheromone likely to locate and/orenter baitedandunbaitedpyr- grandisoic acid (Pin˜ ero et al. 2001). Given that the amidandpanel traps in the presence of host apple principal attractant used in our study, benzaldehyde is trees. a host plant-produced compound, we predicted that The impact of host apple trees couldbe described plum curculio woulddiscriminatebetween baitedand in terms of the types of stimuli they provide foraging unbaitedtraps in the Þeld,but not in the orcharddue plum curculio. For example, many phytophagous in- competition from natural host apple tree odor. Indeed, sects use visual stimuli to aidthem in host plant Þnding we observedno signiÞcant differenceamong baited (Jolivet 1998). In host apple trees, dimensional fea- andunbaitedtraps in any of the orchardsubplots for tures of the trunk andupright limbs (Leskey and either trap type from either sex of plum curculio. We Prokopy 2002) are likely important to plum curculio only recaptured a total of 14 adults in the orchard, 12 because many phytophagous insects respondto tall of which were recapturedin the 5-m subplot. As pre- narrow dimensions of vertically growing plant struc- dicted, we were able to detect discrimination among tures (Prokopy andOwens 1983). This type of visual baitedtraps in our 5-m Þeldsubplot. Here, signiÞ- cue has been usedto create monitoring traps for plum cantly more female plum curculio were recapturedby curculio. The black pyramidtraps usedin our studies pyramidtraps baitedwith benzaldehydecompared originally were designed to capture the pecan weevil with unbaitedpyramidtraps. Pyramidtraps baited andare believedto exploit responses to visual cues with benzaldehyde in combination with grandisoic provided by the tree trunk (Tedders and Wood 1994, acidor with grandisoicacidalone were intermediate Mulder et al. 1997). We hypothesized that if presence in terms of their attractiveness (Table 5). Thus, it of host apple trees interferes with visual stimuli pro- seems possible that plum curculio response to baited videdbypyramidtraps, then recaptures shouldbe traps was diminished due to presence of competition lower in the orchardcomparedwith the Þeld.In our by natural host apple tree odor. Similar results have studies,we observeda nearly ninefolddropin average been reportedfor an attractant available to monitor number of plum curculio recapturedin black pyramid codling moth, Cydia pomonella (L.). The attractant traps placedin our orchard(0.11 Ϯ 0.04 plum curculio ϩ ethyl (2,E,4Z)-2Ð4-decadienoate (a pear-derived vol- per trap) comparedwith the Þeld(0.96 0.13 plum atile) was highly attractive to codling moth in walnut curculio per trap). Given that groundcover was iden- andapple (in the early part of the season) orchards, but was less attractive in pear andapple orchards(in Table 5. Mean number of male and female plum curculio the latter part of the season), owing to the olfactory recaptured by baited pyramid traps in 5-m ﬁeld subplot competition from the same or similar esters likely releasedby maturing or damagedfruit(Light et al. Male Female Bait n Mean Ϯ SEa Mean Ϯ SEb 2001). However, we cannot dismiss an alternative hypoth- Ϯ Ϯ Benzaldehyde 4 1.25 0.25a 4.00 0.71a esis to explain these results, i.e., volatile plumes em- Benzaldehyde ϩ 4 2.00 Ϯ 1.08a 2.25 Ϯ 0.95ab grandisoic acid anating from baitedtraps in the orchardwere simply Grandisoic acid 4 2.00 Ϯ 1.08a 1.50 Ϯ 0.50ab obscured and/or physically disrupted due to the pres- Unbaited4 1.25 Ϯ 0.63a 0.50 Ϯ 0.50b ence of apple trees, creating variation in air movement andmicroenvironment surroundingtraps compared a Values are not signiÞcantly different according to one-way ANOVA (F ϭ 0.27, P ϭ 0.85). with conditions found in the open Þeld. Physical fea- b Values are signiÞcantly different according to one-way ANOVA tures of a particular habitat have inßuencedtrap cap- (F ϭ 4.60, P ϭ 0.023) andTukeyÕs HSD. tures for other weevil species. For example, the boll April 2004 LESKEY AND WRIGHT:INFLUENCE OF HOST TREE 395 weevil, Anthonomus grandis grandis Boheman, is con- potential traps (Tedders and Wood 1994; Mulder et al. sidered to be weak-ßying species. Sappington and 1997; Prokopy andWright 1998; Prokopy et al. 1999, Spurgeon (2000) foundthat variation in dailytrap 2000; Leskey andProkopy 2002), competition from captures was relatedto dailyvariation in synoptic and/or physical obstruction by host trees signiÞcantly windspeed.Furthermore, captures in traps locatedon decreases capture of plum curculio. Deploying baited either sideof a windbreak were inßuencedby the monitoring traps some distance away from the orchard effects produced by the local vegetation structure on itself (Prokopy et al. 2000, Pin˜ ero et al. 2001) may be air ßow. A means to test the hypothesis that host apple one methodfor increasing effective detectionof plum trees simply obscuredor disruptedabilityof plum curculio immigration into orchards but our inability to curculio to locate our baitedmonitoring traps would predict potential injury to fruit based on trap captures be to set up an experiment in a planting of nonhost remains a serious issue. Thus, a secondmethodcould trees with similar architecture andspacing to that be to deploy a larger number of traps to increase foundwithin our experimental orchard.If results ob- chances of plum curculio capture. A thirdmethod servedwithin the orchardwere dueto the physical proposedto overcome weaknesses associatedwith features of the host apple trees themselves, we would monitoring traps is that of a “trap tree” approach that expect to observe identical results in a nonhost tree uses semiochemical cues to aggregate plum curculio planting. This same hypothesis also wouldapply to the andoviposition damageto a particular host tree to potential impact of visual obstruction from host apple make management decisions (Prokopy et al. 2003). As trees. Although we did not test this hypothesis spe- with development of a trap-based curculio monitoring ciÞcally in our studies, other studies have deployed strategy, this approach must reliably overcome com- traps outside orchards along the edge of wood lots, petition from and/or obstruction by unbaited, attrac- where presumably, olfactory andvisual cues pre- tive host trees. Our Þndings suggest that plum curculio sentedby baitedmonitoring traps were at least some- trap captures decrease signiÞcantly when traps are what physically obstructedby nonhost tree species. In locatedwithin an orchardandwe concludethat an these studies, discrimination between baited and un- effective monitoring system for plum curculio will baitedtraps was recorded(Prokopyet al. 2000, Pin ˜ ero require baitedtraps that can compete with host apple et al. 2001), but when baitedtraps were deployed tree stimuli and/or that greater numbers of baited within an orchard, discrimination was not observed traps be deployed to more reliably capture adult plum (Prokopy et al. 2000), indicating that at least to some curculio and predict damage potential, ultimately to degree, particularly at close range (within 1Ð2 m) determineneedandtiming of insecticideapplication. competition from host apple trees must inßuence the ability of plum curculio to locate and/or enter baited monitoring traps. Acknowledgments More importantly, our results highlight an even We thank Henry Hogmire, DavidMandy,andSteve Bliz- greater problemÑour inability to capture mated, ovi- zardfor help in facilitating these studies.We also thank positing females in our monitoring traps to reliably Sridhar Polavarapu and Doug Pfeiffer for helpful comments predict damaging potential of a population. Of those on an earlier draft of this manuscript. females recapturedin our experiments, nearly all were matedwith mature eggs in the calyces. In the Þeld,we were able to recapture 54 females, but only three in References Cited the orchard. Furthermore, we recaptured signiÞcantly Buchbauer, G., L. Jirovetz, J. Wasicky, and A. Nikiforov. more males than females in the orchard(Table 3), but 1993. Headspace and essential oil analysis of apple ßow- not in the Þeld(Table 4). Thus, to effectively monitor ers. J. Agric. FoodChem. 41: 116Ð118. plum curculio and more accurately predict damage Butkewich, S. L., and R. J. Prokopy. 1997. Attraction of adult potential, a baitedmonitoring trap must be capable of plum curculios to host-tree odor and visual stimuli in the attracting andcapturing ovipositing females after fruit Þeld. J. Entomol. Sci. 32: 1Ð6. set, the phenological periodwhen traps lose their Dixon, B. M., R. J. Prokopy, and B. B. Schultz. 1999. Inßu- effectiveness andthe damagepotential of a population ence of weather andtime of dayon plum curculio (Coleoptera: Curculionidae) trap canopy entry behav- increases (Prokopy et al. 2000). Thus, our results ex- iors andevaluation of traps for predictingfruit injury. plain why trap captures recordedforpyramidand/or J. Entomol. Sci. 34: 191Ð202. panel traps deployed in the border rows of orchards Eller, F. J., and R. J. Bartelt. 1996. Grandisoic acid, a male- declinedafterfruit set, failedto reßect amount of produced aggregation pheromone for the plum curculio, oviposition injury observedin fruit trees, andthere- Conotrachelus nenuphar. J. Nat. Prod. 59: 451Ð453. fore failedto serve as reliable tools to determineneed Foster, S. P., and M. O. Harris. 1997. Behavioral manipula- for andtiming of insecticideapplication (Prokopy et tion methods for insect pest management. Annu. Rev. al. 1999, 2000; Leskey andWright 2004). Entomol. 42: 123Ð146. Although progress has been made in terms of iden- Giblin-Davis, R. M., T. J. Weissling, A. C. Oehlschlager, and L. M. Gonzalez. 1994. Fieldresponse of Rhynchophorus tiÞcation of attractive host plant-basedattractants cruentatus (Coleoptera: Curculionidae) to its aggregation (Leskey andProkopy 2000, 2001; Prokopy et al. 2000; pheromone andfermenting plant volatiles. Fla. Entomol. Leskey et al. 2001; Pin˜ ero et al. 2001), identiÞcation of 77: 164Ð177. an attractive component of the male-produced pher- Giblin-Davis, R. M., J. E. Pen˜ a, A. C. Oehlschlager, and omone (Eller andBartelt 1996), anddevelopmentof A. L. Perez. 1996. Optimization of semiochemical-based 396 ENVIRONMENTAL ENTOMOLOGY Vol. 33, no. 2

trapping of Metamasius hemipterus sericeus (Olivier) Prokopy, R. J., and B. A. Croft. 1994. Apple insect pest man- (Coleoptera: Curculionidae). J. Chem. Ecol. 22: 1389Ð agement, pp. 543Ð586. In R. Metcalf andW. Luckman 1410. [eds.], Introduction to insect pest management, 3rd ed. Johnson, D. T., P. G. Mulder, D. McCraw, B. A. Lewis, Wiley, New York. B. Jervis, B. Carroll, and P. J. McLeod. 2002. Trapping Prokopy, R. J., and S. E. Wright. 1998. Plum curculio re- plum curculio Conotrachelus nenuphar (Herbst) (Co- sponses to unbaitedpyramidandcone traps. J. Econ. leoptera: Curculionidae) in the southern United States. Entomol. 91: 225Ð234. Environ. Entomol. 31: 1259Ð1267. Prokopy, R. J., S. S. Cooley, and P. L. Phelan. 1995. Bioassay Jolivet, P. 1998. Interrelationship between insects and approaches to assessing behavioral responses of plum plants. CRC, New York. curculio adults (Coleoptera: Curculionidae) to host fruit Landolt, P. J. 1997. Sex attractant andaggregation phero- odor. J. Chem. Ecol. 21: 1073Ð1084. mones of male phytophagous insects. Am. Entomol. 43: Prokopy, R. J., M. Marsello, T. C. Leskey, and S. E. Wright. 12Ð21. 1999. Evaluation of unbaitedtraps for monitoring and Landolt, P. J., and T. W. Phillips. 1997. Host plant inßuences controlling plum curculio adults (Coleoptera: Curculion- on sex pheromone behavior of phytophagous insects. idae) in apple orchards. J. Entomol. Sci. 34: 144Ð153. Annu. Rev. Entomol. 42: 371Ð391. Prokopy, R. J., B. W. Chandler, T. C. Leskey, and S. E. Wright. Leskey, T. C., and R. J. Prokopy. 2000. Sources of apple odor 2000. Comparison of traps for monitoring plum curculio attractive to adult plum curculios (Coleoptera: Curcu- adults (Coleoptera: Curculionidae) in apple orchards. lionidae). J. Chem. Ecol. 26: 639Ð653. J. Entomol. Sci. 35: 411Ð420. Leskey, T. C., and R. J. Prokopy. 2001. Adult plum curculio Prokopy, R. J., P. L. Phelan, S. E. Wright, A. J. Minalga, (Coleoptera: Curculionidae) attraction to fruit and con- R. Barger, and T. C. Leskey. 2001. Compounds from speciÞc odors. Ann. Entomol. Soc. Am. 94: 275Ð288. host odor attractive to plum curculio adults (Coleoptera: Leskey, T. C., and R. J. Prokopy. 2002. Developing a branch- Curculionidae). J. Entomol. Sci. 36: 122Ð134. mimicking trap for adult plum curculios. Entomol. Exp. Appl. 102: 253Ð259. Prokopy, R. J., B. W. Chandler, S. A. Dynok, and J. Pin˜ ero. Leskey, T. C., and S. E. Wright. (2004). Monitoring plum 2003. Odor-baited trees: a new approach to monitoring curculio populations in apple andpeach orchardsin the plum curculio (Coleoptera: Curculionidae). J. Econ. En- mid-Atlantic. J. Econ. Entomol. 97: 79Ð88. tomol. 96: 826Ð834. Leskey, T. C., R. J. Prokopy, S. E. Wright, P. L. Phelan, and Racette, G., G. Chouinard, C. Vincent, and S. B. Hill. 1992. L. W. Haynes. 2001. Evaluation of individual compo- Ecology andmanagement of plum curculio, Conotrach- nents of plum odor as potential attractants for adult plum elus nenuphar (Coleoptera: Curculionidae) in apple or- curculios. J. Chem. Ecol. 27: 1Ð17. chards. Phytoprotection 73: 85Ð100. Light, D. M., A. L. Knight, C. A. Henrick, D. Rajapaska, Sappington, T. W., and D. W. Spurgeon. 2000. Variation in B. Lingren, J. C. Dickens, K. M. Reynolds, R. G. Buttery, boll weevil captures in pheromone traps (Coleoptera: G. Merrill, J. Roitman, and B. C. Campbell. 2001. A Curculionidae)arising from windspeedmoderationby pear-derived kairomone with pheromonal potency that brush lines. Environ. Entomol. 29: 807Ð814. attracts male andfemale codlingmoth, Cydia pomonella SAS Institute. 2001. Version 8.2, SAS Institute, Cary, NC. (L.). Naturwissenschaften 88: 333Ð338. Smith, E. H., and E. H. Salkeld. 1964. Ovary development Maier, C. T. 1990. Native andexotic rosaceous hosts of andoviposition rates in the plum curculio, Conotrachelus apple, plum, andquince curculio larvae (Coleoptera: nenuphar (Coleoptera: Curculionidae). Ann. Entomol. Curculionidae) in northeastern United States. J. Econ. Soc. Am. 57: 781Ð787. Entomol. 83: 1326Ð1332. Tedders, W. L., and B. W. Wood. 1994. A new technique for Mulder, P. G., B. D. McCraw, W. Reid, and R. A. Grantham. monitoring pecan weevil emergence (Coleoptera: 1997. Monitoring adult weevil populations in pecan and Curculionidae). J. Entomol. Sci. 29: 18Ð30. fruit trees in Oklahoma. Okla. State Univ. Ext. Facts F. Thomson, R. R. 1932. Sex differentiation of adults of Cono- 7190: 1Ð8. trachelus nenuphar. J. Econ. Entomol. 25: 807Ð810. Oehlschlager, A. C., C. M. Chinchilla, L. M. Gonzalez, Vincent, C., G. Chouinard, and S. B. Hill. 1999. Progress in F. Jiron, R. Mexzon, and B. Morgan. 1993. Development plum curculio management: a review. Agric. Ecosyst. of a pheromone-basedtrapping system for Rhynchopho- Environ. 73: 167Ð175. rus palmarum (Coleoptera: Curculionidae). J. Econ. En- Yonce, C. E., D. L. Horton, and W. R. Okie. 1995. Spring tomol. 86: 1381Ð1392. migration, reproductive behavior, monitoring proce- Pin˜ ero, J. C., S. E. Wright, and R. J. Prokopy. 2001. Response dures, and host preference of plum curculio (Coleoptera: of plum curculio (Coleoptera: Curculionidae) to odor- Curculionidae) on Prunus species in Central Georgia. baitedtraps near woods.J. Econ. Entomol. 94: 1386Ð1397. J. Entomol. Sci. 30: 82Ð92. Prokopy, R. J., and E. D. Owens. 1983. Visual detection of plants by herbivorous insects. Annu. Rev. Entomol. 28: 337Ð364. Received 14 July 2003; accepted 23 December 2003.