H E Use of Residue Proﬁle Analysis to Identify Modes of ...Gut 2000). Although such Þeld trials are encouraging in terms of showing basic fruit protection over pre-scribed intervals,

HORTICULTURAL ENTOMOLOGY

Use of Residue Profile Analysis to Identify Modes of InsecticideActivity Contributing to Control of Plum Curculio in Apples

JOHN C. WISE,1,2 ANDREA B. COOMBS,1 CHRISTINE VANDERVOORT,3 LARRY J. GUT,1

ERIC J. HOFFMANN,1 AND MARK E. WHALON1

J. Econ. Entomol. 99(6): 2055Ð2064 (2006)

ABSTRACT Residue proÞle analysis techniques were developed, along with laboratory and Þeld-based bioassays to describe the modes of insecticidal activity responsible for the control of the plumcurculio,Conotrachelus nenuphar (Herbst), in apples (Malus spp.). Adult plum curculios were treatedin laboratory topical bioassays to determine acute contact activity and lethal time for Þve insecticides.Azinphosmethyl had the highest levels of toxicity and shortest lethal time values, followed by theneonicotinoids thiacloprid, thiamethoxam, and imidacloprid, whereas indoxacarb had the highestLD50 and LT50 values for topical exposure. Field-based residual activity bioassays assessed adultmortality, and fruit and leaf injury from plum curculio exposed to 4 h, 7 d, and 14 d Þeld-aged residues.All compounds caused signiÞcant levels of mortality to plum curculio when adults were exposed tofruit clusters 4 h post-application. Thiacloprid, thiamethoxam, and imidacloprid showed ovipositiondeterrence, antifeedant, and repellency effects in the 7- and/or 14 d residual bioassays and protectedfruit in the absence of signiÞcant lethal activity. Indoxacarb maintained lethal activity throughout thestudy intervals, with the incidence of plum curculio feeding, suggesting that ingestion is an importantmode of entry. For the neonicotinoids thiacloprid, thiamethoxam and imidacloprid plum curculiomortality was highly correlated with fruit and leaf surface residues. As surface residues declined,sublethal effects such as oviposition deterrence and antifeedant effect remained. The value of theplantÐinsectÐchemistry triad model for describing the temporal dimensions of insecticidal modes ofactivity and understanding a compoundÕs critical performance characteristics is discussed.

KEY WORDS Conotrachelus nenuphar, apple, residue proÞle, plantÐinsectÐchemistry triad, sub-lethal effects

The plum curculio, Conotrachelus nenuphar (Herbst),is a native beetle that occurs east of the Rocky Moun-tains (Chapman 1938) and is a key pest of cultivatedpome and stone fruits in eastern and central NorthAmerica (Racette et al. 1992, Yonce et al. 1995). In thespring, female plum curculios make crescent-shapedoviposition scars in the developing fruit, and larvaefeed internally on the fruit ßesh. Adult plum curculiosalso damage fruit by making a small puncture hole inthe skin and then eating the ßesh underneath the skin.Internal feeding damage can render the fruit unmar-ketable, as does the scar tissue resulting from ovipo-sition and adult feeding. Developing larvae tunnel inthe fruit and have been reported to cause fruit ab-scission (Levine and Hall 1977); however, it is difÞcultto determine effects on yield in apple because curcu-lio-induced fruit abscission occurs at the same time asJune drop. Left unchecked, the plum curculio cancause signiÞcant damage in apple (Malus spp.) or-chards in a short time. In early studies of plum curculio

damage potential in apple, Stedman (1904) found thata single female curculio can lay between 200Ð400 eggsand make about twice as many oviposition scars.Prokopy et al. (1993) found that plum curculio cancause economic levels of injury in a 12-h period; con-trol measures should be well timed and broadly ef-fective to protect the crop.

Over the past 50 yr, plum curculio control has beenbased almost entirely on organophosphate insecti-cides, primarily azinphosmethyl (Guthion, BayerCropScience, Research Triangle Park, NC) and to alesser extent phosmet (Imidan, Gowan Company,Yuma, AZ). In recent years, Michigan apple growershave effectively prevented plum curculio injury withtwo to three applications of organophosphate insec-ticides. The effectiveness and availability of organo-phosphates (OPs) have resulted in limited perceivedneed for research on alternative control measures.However, as a result of the Food Quality ProtectionAct (FQPA) (FQPA 1996), apples grown east of theMississippi are now limited to a total of 1.8 kg (4.0 lb)active ingredient (AI)/acre of azinphosmethyl peryear, leaving growers with three to four applications tocover a multitude of key pests throughout the season.Continued use of azinphosmethyl is subject to regu-

1 Department of Entomology, Michigan State University, 243 Nat-ural Science, East Lansing MI 48824.

2 Corresponding author, e-mail: [email protected] Michigan State University Pesticide Analytical Laboratory, 206

Center For Integrated Plant Systems, East Lansing MI 48824.

0022-0493/06/2055Ð2064$04.00/0 � 2006 Entomological Society of America

latory review, and this product may in the futurebecome unavailable to growers.

Since the implementation of FQPA there have beena series of new reduced-risk and OP-alternative in-secticides that show promise for plum curculio con-trol. The neonicotinoid compounds Actara (thia-methoxam), Calypso (thiacloprid), and oxidiazineAvaunt (indoxacarb) have all shown signiÞcant levelsof control in Þeld efÞcacy trials on apple (Wise andGut 2000). Although such Þeld trials are encouragingin terms of showing basic fruit protection over pre-scribed intervals, they do not provide much informa-tion on the basis of the insecticideÕs performance.Optimizing the performance of a compound requiresunderstanding the mode of exposure, Þeld residual,and modes of insecticidal activity that are responsiblefor preventing the pest from injuring the crop. Recentresearch on compounds in the neonicotinoid insecti-cide class suggests that sublethal modes of activity(i.e., antifeedant, oviposition deterrence) can make animportant contribution to their overall performance(Hu and Prokopy 1998, Kunkel et al. 2001, Liburd etal. 2003, Isaacs et al. 1999, Nauen 1995). In addition,neonicotinoid insecticides are known to have uniquechemodynamic interactions with the plant, which caninßuence insecticide exposure to the pest (Mac-Donald and Meyer 1998, Wamhoff and Schneider1999). The results of these studies suggest that thespatial and temporal dimensions of the plantÐinsectÐchemical (PIC) interaction may be important in fullyunderstanding the overall performance of a pest man-agement tool.

In this study, we combined residue proÞle analysiswith Þeld and laboratory bioassays to describe themodes of insecticidal activity that are responsible forthe control of the plum curculio. The four organo-phosphate-alternative insecticides included in thestudy were the neonicotinoids imidacloprid, thiaclo-prid, and thiamethoxam and the oxadiazine indox-acarb, with the organophosphate azinphosmethyl asthe standard comparison. The objectives of this studywere to 1) characterize the lethal activity of thesecompounds in terms of lethal dose and lethal time, 2)identify the nonlethal modes of insecticidal activitythat contribute to performance and the temporal di-mension of each, and 3) describe the relationshipbetween surface and interior residues on the plant andthe lethal activity of insecticides on the plum curculio.

Materials and Methods

Insect Material. Northern Strain. For trials evaluat-ing the behavior of plum curculio, northern (diapaus-ing) strain insects were collected in May and June2001 with beating trays and pyramid traps (Teddersand Wood 1994) placed in orchards at the MichiganState University Trevor Nichols Research Complex.Beetles were collected 4Ð7 d before trial use and heldtogether at 25�C at a photoperiod of 16:8 (L:D) h toprovide time for mating. Females were sorted frommales 24 h before study initiation.

Southern Strain. To improve consistency of results,beetle age, and history were standardized for bioas-says. For these trials, southern (nondiapausing) strainweevils were used from a continuous colony at Mich-igan State University. This colony has been in pro-duction since 1998 with occasional additions from col-laborators from the southeastern United States. Adultswere reared on immature apples for food and ovipo-sition substrate (modiÞed from Smith 1957) and heldat 25�C at a photoperiod of 16:8 (L:D) h. Larvae wereplaced in soil for pupation, and adults were collectedin traps above these pupation containers.Laboratory Topical Toxicity Bioassay. Topical bio-

assays were used to determine the baseline toxicity ofimidacloprid (Bayer CropScience, Research TrianglePark, NC), thiacloprid (Bayer CropScience), thia-methoxam (Syngenta, Greensboro, NC), indoxacarb(DuPont, Wilmington, DE), and azinphosmethyl(Bayer CropScience) to plum curculio adults. Tech-nical grade material was serially diluted with acetone(98.7% purity) to eight concentrations (31,600, 10,000,3,160, 1,000, 316, 100, 31.6, and 10 �g [AI]/ml). Adultsouthern strain beetles were used (age 1Ð2 wk post-eclosion). Beetles were treated with 1 �l of eachsolution on the dorsal surface of the abdomen appliedwith repeating dispenser (Hamilton PB600-1)equipped with a 50-�l beveled point syringe (Hamil-ton,Reno,NV).Beetles in thecheckwere treatedwithacetone only. The actual doses of active ingredientapplied to beetles were 31.6, 10, 3.16, 1, 0.316, 0.1,0.0316, and 0.01 �g (AI) per beetle. After treatment,beetles were placed in a 100- by 15-mm polystyrenedisposable sterile petri dish (VWR, West Chester, PA)lined with a Whatman no. 2 Þlter paper (Maidstone,England). Immature ÔEmpireÕ apples (�25 mm in di-ameter) were sliced into 10-mm-thick sections placedinside each of the petri dishes as a food source. Appleswere replaced daily. Petri plates were randomly la-beled for blind readings. Mortality was recorded at 3,6, 12, 24, 48, 72, and 120 h after application to deter-mine lethal dose and lethal time aspects of toxicity.Each replicate contained 10 beetles, with three rep-lications per treatment.

Mortality data were corrected for natural mortalityby using AbbottÕs formula (Abbott 1925). The LC50,Þducial limits, slope � SE, and chi-square goodness-of-Þt values were determined using probit analysis(SAS Institute 2002). Relative toxicity of each com-pound to azinphosmethyl was calculated by dividingthe LD50 value of the compound by that of azinphos-methyl. Percentage of mortality data also were arc-sine-transformed for analysis of variance (ANOVA)(Sokal and Rohlf 1994).Field Trials. Plots consisted of two 9-yr-old ÔGalaÕ

apple,Malus domestica Borkhausen, trees. The exper-imental design was a randomized complete block withfour replications. Tree spacing was 5.5 by 6.1 m, withone buffer tree and one buffer row separating all plots.Regular maintenance applications of fungicides wereapplied to all treatment blocks. Insecticide treatmentswere applied with an FMC 1029 airblast sprayer cal-ibrated to deliver 935 liters/ha (100 gpa). Applications

2056 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 99, no. 6

were made on 17 May (petal fall � 3 d). A secondapplication was made on 31 May for the efÞcacy trialassessment. Fruit and foliage from one tree persprayed and untreated check plot were collected foruse in residue proÞle analysis and the Þeld-based bio-assays. Fruit on the second tree of each plot were usedfor the efÞcacy trial.

Treatmentcompounds, formulations, andratesusedfor the applications were imidacloprid (Provado 1.6 F,Bayer CropScience) 111.83 g (AI)/ha (8 ß oz/acre),thiacloprid (Calypso 4 F, Bayer CropScience) 105.08g (AI)/ha (3 ß oz/acre), thiamethoxam (Actara 30WG, Syngenta) 94.57 g (AI)/ha (4.5 oz/acre), andindoxacarb (Avaunt 30 WG, DuPont) 126.09 g(AI)/ha (6.0 oz/acre). A nonionic spreader sticker,Kinetic (Helena Chemical Co., Collierville, TN) wasadded to the Avaunt formulation per manufacturerÕsrecommendation. The organophosphate azinphos-methyl (Guthion 50W, Bayer CropScience) 1,120.84 g(AI)/ha (2 lb/acre) was added to this set treatmentsfor the Þeld efÞcacy trial and bioassays related toresidue proÞle analysis.FieldEfficacyTrial. As a measure of the overall Þeld

performance of insecticides, apples were evaluated foroviposition damage on 15 June 2001 (15 d after Þnaltreatment application). For evaluation, fruit were ran-domly harvested (100 fruit per plot, 400 fruit pertreatment), and the number of fruit with ovipositionstings was recorded. Fruit injury percentages werearcsine square-root transformed and compared usingANOVA (Kerchove 2005).Residual Activity Bioassays. Field-based bioassays

were used to evaluate the modes of insecticidal ac-tivity that contribute to performance and the temporalprogression of those modes as treatment residues age.Fruit clusters were collected from Þeld treated trees4 h after the applications and again 7 and 14 d later.Each shoot was pruned to have 10 fruit and 10 leavesand placed in water-soaked OASIS ßoral foam (Smith-ers-Oasis Co., Kent, OH) in clear plastic 946-ml con-tainers (Fabri-Kal, Kalamazoo, MI) with lids. Thefoam was covered with sealing wax (Gulf Wax, dis-tributed by Royal Oak Sales, Inc, Roswell, GA) topreserve the integrity of the fruit and foliage. Holeswere punched in each containerÕs lid to reduce con-densation of water vapor inside the container andminimize fumigation effects, if any. Each of thesecontainers was considered an experimental unit in thebioassay.

Northern strain adults were collected, held, andsexed 24Ð48 h before each bioassay. Five females wereplaced in the bottom of each experimental arena.There were four replicates for each treatment at eachof the post-application time intervals. The numbers ofliving and moribund beetles were recorded after 96 hof exposure. The numbers of fruit oviposition stings,fruit feeding marks, and leaf-feeding marks also wererecorded. Mortality and/or moribund was deÞned asbeetles that were dead or rendered unable to recoveror function normally after being held over a 24-hperiod. Beetles were counted as alive if they seemedto function normally. Statistical comparisons (by

date) were made using ANOVA (PROC GLM, SASInstitute 2002). Incremental behavior observationswere made on beetles after 1, 6, 12, 24 and 96 h ofexposure in the bioassay arenas for each of the threeresidual bioassays (4 h, 7 d, and 14 d postinsecticideapplication). Beetles were observed for 1 min at eachtime interval, and data were recorded on the positionand/or activity of beetles, categorized as follows: bee-tle on leaf/stem, beetle feeding on fruit, beetle ovi-positing in fruit, beetle still on fruit, beetle alive butnot on plant, beetle twitching on bottom, and beetledead/still on bottom. “Still” beetles were not manip-ulated to determine mortality so as to not disturb theongoing experiment. Statistical comparisons weremade for each treatment to the untreated check byusing DunnettÕs test (Kerchove 2005).Residue Profile Analysis.Residue Profiles.A parallel

series of fruit and foliage samples were taken fromÞeld plots at each of the three postapplication timings(4 h, 7 d, and 14 d). Composite samples of 40 leaves andfruit each were frozen and transported to the Mich-igan State University pesticide analytical laboratory inEast Lansing, MI. A novel pesticide extraction proce-dure was used to separate dislogeable residues on thesurface of the fruit and leaf samples from the propor-tions in and below the plant cuticle, to provide a spatialproÞle of each compound over time.

To determine the amount of residue on the fruit andleaf surfaces, 10-g composite samples of apple fruit andleaves were placed in 150 ml of acetonitrile and son-icated for 10Ð15 s. The acetonitrile was decantedthrough 5 g of anhydrous sodium sulfate to remove allwater. The sample was dried via rotary evaporationand brought up in acetonitrile for high-performanceliquid chromatography (HPLC) or gas chromatogra-phy/mass spectrometry (GC/MS) analysis.

To determine the residue from the fruit and leafinteriors, the remaining solid fruit and leaf sampleswere ground with 200 ml of dichloromethane. Theextracts were then vacuum-Þltered, and the Þltratewas passed through 5 g of anhydrous sodium sulfate.The samples were dried via rotary evaporation andbrought up in acetonitrile. Any remaining particulateswere removed by passing the sample through a0.45-�m Þlter.

Thiamethoxam, thiacloprid, and imidacloprid sam-ples were analyzed for insecticide residue with a Wa-ters 2690 separator module HPLC equipped with aWaters 2487 dual-wavelength absorbance detector setat 254, 242, or 270 nm, respectively, and a C18 reversedphase column (4.6-mm bore, 5-mm particle size). Themobile phase was water/acetonitrile (80:20) at 55�C.Indoxacarb and azinphosmethyl were analyzed usingGC/MS. All compounds were quantitated against astandard curve, Þtted linearly with a range typicallybetween 2.5 and 20 �g/g. A 10-�l injection was usedin HPLC analysis and a 1-�l injection with the GC/MS.Residue Correlation Bioassays. Southern strain bee-

tles were used to assess the relationship between theÞeld-aged residues and the plum curculio mortalityÐresponse because of the ability to control for uniform

December 2006 WISE ET AL.: PLUM CURCULIO AND MODES OF INSECTICIDE ACTIVITY 2057

age and physiological condition in a laboratory colony.Females used in bioassays were 1Ð2 wk post-eclosionand were exposed to males to assure mated status. Tenfemale beetles were used in each experimental unit,with four replicates per treatment at each of the threepostapplication time intervals (4 h, 7 d, and 14 d).Beetles were examined after 96 h of exposure to Þeldtreated fruit clusters (see bioassay container descrip-tion above), and a Þnal count was made of moribundbeetles, along with the amount of leaf feeding andoviposition damage. For each treatment compound,regression (PROC GLM, SAS Institute 2002) was usedto determine correlates between curculio mortalityand type of residue (surface and interior leaf residues,surface and interior fruit residues).

Results

Laboratory Topical Toxicity Bioassay. Azinphos-methyl was the most toxic compound to plum curculioadults in the topical assays, with an LD50 value of 0.16�g per beetle (160 ppm) (Table 1). Thiacloprid andthiamethoxam were less toxic, but their 95% conÞ-dence intervals overlapped with azinphosmethyl. Imi-dacloprid followed thiacloprid and thiamethaxam intoxicity, and indoxacarb was the least toxic of theinsecticides tested in the topical bioassays.

Time to 50% mortality (LT50) is shown in Table 2.Guthion had the most rapid lethal activity with anLT50 value of 0.76 h at 10 �g (AI) per beetle (highestdose). Imidacloprid had lethal activity with a calcu-lated LT50 value of 18.27 h at 10 �g (AI) per beetledose. Thiamethoxam and thiacloprid had LT50 valuesof 39.25 and 43.99 h, respectively, at 10 �g (AI) perbeetle. Indoxacarb showed the slowest activity havingan LD50 value of 114 h at 10 �g (AI) per beetle.Field Trials. Field Efficacy Trial. Untreated check

trees had an average of 43.0% of fruit damaged fromplum curculio oviposition. All treatments except imi-

dacloprid provided signiÞcant levels of fruit protec-tion compared with the untreated check (F � 4.25,df � 18, P � 0.01) (Table 3). Imidacloprid-treatedtrees had signiÞcantly more damage than thiaclopridand thiamethoxam but not signiÞcantly different fromthe other two compounds.Residual Activity Bioassays. Azinphosmethyl could

not be included in the treatment list because of thelimited numbers of northern strain beetles collectedfrom the Þeld. Plum curculio adults exposed to 4 hpost-application chemical residues resulted in signif-icantly higher levels of mortality for all treatmentcompounds than the untreated check (F� 3.16, df �15, P � 0.045) (Table 4). The incidence of fruit ovi-position and fruit and leaf feeding was limited signif-icantly for all compounds compared with the un-treated check (oviposition: F � 66.794, df � 15, P �0.0001; fruit feeding: F � 31.304, df � 15, P � 0.0001;and leaf feeding: F � 25.733, df � 15, P � 0.0001). Infruit feeding, thiacloprid, thiamethoxam, and indox-acarb provided better fruit protection than imidaclo-prid.

For plum curculio adults exposed to 7-d-old resi-dues, only indoxacarb caused a signiÞcant level ofmortality (F � 12.84, df � 15, P � 0.0001). The inci-dence of fruit oviposition and fruit feeding, however,was signiÞcantly reduced for all compounds comparedwith the untreated check (oviposition: F� 10.06, df �15,P� 0.0004; and fruit feeding:F� 7.929, df � 15,P�0.0012). In fruit oviposition, thiacloprid and indox-acarb maintained the lowest levels, followed by imi-dacloprid and thiamethoxam. For the neonicotinoidcompounds, the reduced feeding and oviposition inthe presence of predominantly live beetles in thecanopy suggests oviposition deterrence and antifeed-ant modes of activity at work.

For plum curculio adults exposed to 14-d-old resi-dues, only indoxacarb caused a signiÞcant level ofbeetle mortality (F� 8.06, df � 15,P� 0.0011). By day

Table 1. Probit analysis (lethal dose) after 120 h for five chemicals applied to southern strain plum curculio

Compound n Slope � SE LD50 (95% FL) �2 Toxicity ratioa

Azinphosmethyl 30 2.89 � 0.83 0.16 (0.06Ð0.53) �0.001Imidacloprid 30 0.91 � 0.13 3.67 (2.05Ð8.31) �0.001 0.043Thiacloprid 30 0.92 � 0.11 0.285 (0.17Ð0.46) �0.001 0.561Thiamethoxam 30 1.30 � 0.23 0.68 (0.28Ð1.85) �0.001 0.235Indoxacarb 30 1.48 � 0.26 21.7 (13.8Ð43.7) �0.001 0.007

a Toxicity relative to azinphosmethyl.

Table 2. Probit analysis results for lethal time (hours) at 10 �g �AI� per beetle

Compound n Slope � SE LT50 (95% FL) �2 Toxicity ratioa

Azinphosmethyl 30 1.98 � 0.95 0.76 (9.49 1014Ð1.95) 0.03Imidacloprid 30 0.85 � 0.17 18.27 (10.55Ð29.80) �0.001 0.041Thiacloprid 30 2.37 � 0.46 43.99 (27.18Ð78.25) �0.001 0.017Thiamethoxam 30 1.41 � 0.39 39.25 (16.43Ð208) �0.001 0.019Indoxacarb 30 9.07 � 2.18 114 (103Ð132) �0.001 0.006Indoxacarbb 30 7.63 � 1.57 107 (96Ð124) �0.001 0.007

FL, Þducial limit.a Toxicity relative to azinphosmethyl.b 31.6 � g (AI) per beetle.


14, under the no-choice bioassay conditions, therewere no signiÞcant reductions in oviposition injury forany treatments (F � 2.286, df � 15, P � 0.1082).Thiacloprid and indoxacarb signiÞcantly reduced leaffeeding, whereas thiamethoxam reduced fruit feedingcompared with the untreated check (leaf feeding: F�3.92, df � 15, P� 0.0225; and fruit feeding: F� 4.058,df � 15, P � 0.02).

Behavioral observations of northern strain plumcurculio adults in the 4-h Þeld-aged bioassays (eval-uations made after 1, 6, 12, 24, and 96 h of beetleexposure) resulted in documented patterns of behav-ior (and/or canopy positions) that were abnormal incontrast to the beetles in the untreated check (Fig. 1).The most pronounced abnormal behavior occurrencethroughout the Þve observation periods of the 4-hbioassay setup was that of beetles being dead/still onbottom of the bioassay chamber. This condition wassigniÞcantlydifferent fromthecheck for thiacloprid inall Þve observations periods, for imidacloprid and thia-methoxam in four of the Þve periods, and for indox-acarb for two of the periods (1-h exposure: F � 2.55,df � 15, P � 0.0824; 6-h exposure: F � 7.69, df � 15,P � 0.0014; 12-h exposure: F � 11.00, df � 15, P �0.0002; 24-h exposure: F � 12.27, df � 15, P � 0.0001;and 96-h exposure: F� 4.49, df � 15, P� 0.013). Thebeetles in the three neonicotinoid bioassays reachedtheir mortality midpoint after 6Ð12 h of exposure,whereas beetle mortality in the indoxacarb treatmentprogressed very slowly until the 96-h observation(particularly when including the prelethal “twitching”condition included in the moribund measures). ThemortalityÐexposure patterns observed in these Þeld-based bioassays are relatively consistent with the LT50

values produced from the topical bioassay study (Ta-ble 2), reßecting the speed of activity differencesbetween the compounds. There were only two occa-sions of signiÞcantly higher levels of beetles twitchingon bottom, one occasion being in the thiacloprid treat-ment at the 1-h observation and the other occasionbeing in the indoxacarb treatment at the 96-h obser-vation (1-h exposure: F� 4.25, df � 15, P� 0.017; and96-h exposure: F � 30.30, df � 15, P � 0.0001). Gen-erally, the beetles “twitching on bottom” eventuallydied, although for imidacloprid and thiacloprid in the

Table 3. Percentage of fruit with oviposition injury from plumcurculio in a field efficacy trial

TreatmentaProduct rate(g �AI�/ha)

% fruit injury(�SD)b

Untreated check 43.0 (14.7)aAzinphosmethyl 1,120.84 7.0 (3.1)bcImidacloprid 111.83 26.0 (7.8)abThiacloprid 105.08 5.5 (1.7)cThiamethoxam 94.57 4.5 (2.6)cIndoxacarb 126.09 8.5 (5.4)bc

Means followed by the same letter are not signiÞcantly different(P � 0.05; least signiÞcant difference �LSD�).a Insecticide applications made on 17 and 31 May, delivered with

935 liters/ha water.b Raw data square root transformed �SQR(X�0.5)� before ANOVA.

Untransformed means shown for comparison.

Tab

le4

.M

ean

num

bers

(�SE

)of

plan

tti

ssue

dam

agea

and

mor

ibun

dno

rthe

rnst

rain

plum

curc

ulio

afte

rex

posi

ngfiv

efe

mal

esto

field

-age

din

sect

icid

ere

sidu

es

Tre

atm

en

t

Resi

due,4

hpost

-applica

tion

Resi

due,7

dpost

-applica

tion

Resi

due,14

dpost

-applica

tion

Fru

itfe

edin

gL

eaf

feedin

gb

Ovip

osi

tion

stin

gsb

%m

ori

bun

dc

Fru

itfe

edin

gb

Leaf

feedin

gb

Ovip

osi

tion

stin

gsb

%m

ori

bun

dc

Fru

itfe

edin

gb

Leaf

feedin

gb

Ovip

osi

tion

stin

gsb

%m

ori

bun

dc

Un

treat

ed

check

85.8

(14.

6)a

22(3

.2)a

55.5

(5.7

)a0b

52(6

.5)a

0.5

(0.3

)a87

.5(6

.9)a

0b34

.8(1

2.7)

a8.

5(3

.5)a

97.3

(16.

7)a

0bIm

idac

lopri

d16

(7.4

)b2.

5(2

.2)b

4.0

(1.4

)b45

(18.

9)a

18.8

(1.0

)b3.

0(1

.2)a

52.5

(8.1

)bc

0b29

.8(5

.7)a

6.8

(2.0

)ab

47.8

(5.0

)a0b

Th

iacl

opri

d2.

8(1

.4)c

0.8

(0.5

)b1.

5(0

.5)b

65(2

3.6)

a18

(2.8

)b2.

8(1

.5)a

28.5

(8.8

)d10

(5.8

)b32

(5.9

)a1.

0(0

.7)c

61(1

1.5)

a0b

Th

iam

eth

oxa

m0.

3(0

.3)c

0b1.

3(0

.6)b

60(1

8.3)

a22

(7.3

)b1.

0(1

.0)a

59.3

(6.4

)b0b

7.8

(3.8

)b8.

0(2

.9)a

64(1

5.2)

a0b

Indoxa

carb

2.5

(1.0

)c1.

0(0

.4)b

2.0

(0.7

)b70

(12.

9)a

24(2

.1)b

0.8

(0.5

)a34

(3.2

)cd

70(1

9.1)

a39

.5(7

.1)a

1.8

(0.9

)bc

65.8

(8.4

)a25

(9.6

)a

Beetl

es

were

exp

ose

dto

treat

ed

fruit

clust

ers

eac

hw

ith

10fr

uit

and

10le

aves

for

96h

.D

ata

inco

lum

ns

follow

ed

by

the

sam

ele

tter

are

not

sign

iÞca

ntl

ydif

fere

nt

(P�

0.05

).a

Leaf

feedin

gdam

age

iseven

ts/1

0le

aves;

fruit

feedin

gan

dovip

osi

tion

dam

age

are

even

ts/1

0fr

uit

.b

Obse

rved

dat

aw

ere

(squar

ero

ot

�0.

5)tr

ansf

orm

ed

befo

reA

NO

VA

.M

ean

separ

atio

ns

calc

ula

ted

usi

ng

LSD

.D

ata

show

nar

eun

tran

sform

ed

mean

coun

ts.

cO

bse

rved

dat

aw

ere

(arc

sin

esq

uar

e-r

oot

perc

en

t)tr

ansf

orm

ed

befo

reA

NO

VA

.M

ean

separ

atio

ns

calc

ula

ted

usi

ng

LSD

.D

ata

show

nar

eun

tran

sform

ed

mean

s.


6-h exposure observation there seemed to be beetlerecovery before the 12-h period. There were fewerbeetles observed ovipositing in fruit at the 6- and 12-hobservations for all compounds, except not imidaclo-prid at the 12-h period (6-h exposure: F� 112.39, df �15, P � 0.0001; and 12-h exposure: F � 4.47, df � 15,P � 0.014). Similarly there was a lower incidence ofbeetles observed feeding on fruit at the 1- and 24-hperiods for all compounds (1-h exposure: F � 7.56,df � 15, P � 0.0015; 24-h exposure: F � 6.75, df � 15,P � 0.0026).

Behavioral observations for the 7-d bioassay setupdocumented some noteworthy changes in behaviorand/or canopy position patterns (Fig. 1). The mostdramatic pattern change from the 4-h bioassay was inthe incidence of beetles with a dead/still condition.There was little or no lethal response to exposure toimidacloprid, thiacloprid, and thiamethoxam in the7-d bioassay setup. Even so, on two occasions (1- and6-h observation periods), there were signiÞcantlyfewer beetles observed ovipositing in fruit in the neo-nicotinoid treatments (1-h exposure: F� 5.45, df � 15,P � 0.0065; and 6-h exposure: F � 4.25, df � 15, P �0.017). This reduction in oviposition in the absence ofmortality is evidence of an oviposition deterrencemode of insecticidal activity. In addition, at the 6-hobservation for all three neonicotinoids, there weresigniÞcantly higher numbers of beetles alive but noton plant, indicating a repellency mode of activity forthose compounds (6-h exposure: F� 9.46, df � 15, P�0.0005). Indoxacarb was also the only treatment whereexposure resulted in signiÞcantly fewer live beetles onthe plant, compared with the untreated check (96-h

exposure: F � 2.31, df � 15, P � 0.10). The twitchingon bottom condition was also signiÞcantly differentthan the control in the indoxacarb treatment, ob-served at the 96-h period (96-h exposure: F � 22.65,df � 15, P � 0.0001). In addition, in the indoxacarbtreatment there were higher numbers of beetles feed-ing on fruit observed, suggesting that ingestion may bean important contributor to its lethal activity (12-hexposure: F � 5.20, df � 15, P � 0.0078).

Behavioral observations for the 14-d Þeld-aged bio-assays documented far fewer instances of abnormalbehaviors and/or canopy position patterns. At the 1-and 6-h periods for all insecticide treatments therewere still fewer beetles observed ovipositing in fruitthan in the untreated check (1-h exposure: F� 14.49,df � 15, P� 0.0001; and 6-h exposure: F� 9.0, df � 15,P � 0.0006). At the 6-h observation thiacloprid andthiamethoxam treatments also had signiÞcantly lowerincidences of beetles feeding on fruit, indicating someantifeedant activity (6-h exposure:F� 6.0, df � 15,P�0.0003). Imidacloprid again showed a higher numberof beetles alive but not on plant, suggesting repellencyactivity (96-h exposure: F� 6.75, df � 15, P� 0.0026).Residue Profile Analysis. The residue proÞle anal-

ysis shows azinphosmethyl to have predominantly sur-face residues on fruit and leaves, with total activeingredient declining over time (Fig. 2). The moredramatic decline in fruit residues between the 4-h and7-d samples compared with leaf residues is probablyrelated to the growth dilution effect of the rapidlyexpanding fruit tissue at this time of the growing sea-son (Borchert et al. 2004). The residue proÞles ofimidacloprid and thiamethoxam are similar in that

Fig. 1. Percentage of northern strain plum curculio adults observed in behavior or canopy positions at incrementalpostexposure observations (1, 6, 12, 24, and 96 h) for 4 h (A) and 7 d (B) posttreatment residual activity bioassays. The circledstar symbol within individual behavior/position bars for each treatment compound indicate statistically higher incidencecompared with the untreated check (DunnettÕs test), whereas the place within untreated behavior/position bars indicatessigniÞcantly lower incidence within one or more treatments (see text).


large proportions of their active ingredient move intofruit and leaf tissue within 4 h of the Þeld application,and surface residues diminished within the Þrst 7 d ofÞeld exposure. Thiacloprid was seen to have morepersistent surface residues on fruit and leaves, withsome penetration observed into fruit. Indoxacarb wasseen to have relatively equal proportions of activeingredient on the leaf surface as in or below the leafcuticle, and the residue pattern remained stable overtime. On fruit, there were higher proportions of in-doxacarb on the fruit surface than within the fruit.Residue Correlation Bioassays. The regression r2

correlations of fruit (surface, internal) and leaf (sur-face, internal) residues and plum curculio adultmortality (southern strain beetles) showed severalsigniÞcant relationships. For azinphosmethyl thestrongest signiÞcant relationship was between sur-face leaf residue and mortality (R2 � 0.8379, modelsigniÞcant at � � 0.001) (Table 5). This suggests thatthe contact exposure of plum curculio adults to azin-phosmethyl leaf surface residues in the tree canopy isthe most important contributor to mortality as theprimary mode of activity for this compound. For thia-methoxam and thiacloprid, there were strong signiÞ-cant relationships between mortality and leaf and fruitsurface residues (r2 values �0.70, model signiÞcant at� � 0.001) as well as internal leaf residue for thia-methoxam. Imidacloprid showed a similar pattern ofcorrelation for surface and internal leaf residue andsurface fruit, except that the relationships betweenresidue and mortality were generally weaker (r2

values �0.60, model signiÞcant at � � 0.05). Thissuggests that plum curculio contact exposure to theneonicotinoid surface residues on leaf and fruit areresponsible for the important lethal mode of activityseen in the Þrst week after application. In the originalindoxacarb data set, no correlations were found to besigniÞcant for any of the plant residue proÞles. Therewas one mortality data rep in the bioassay that isstrongly suspected to be an outlier, but no clear ex-planation for its occurrence can be determined. Whenthat data rep is eliminated and the data set reanalyzed,several noteworthy correlation relationships emerge.With the corrected indoxacarb data there were sig-niÞcant relationships with mortality for surface andinternal leaf and fruit residues (r2 values �0.50, modelsigniÞcant at � � 0.05), the internal residues beinggenerally stronger than the surface residues. The sig-niÞcant relationship between internal residues andmortality suggests that exposure through ingestionmay be an equally important contributor to its lethalactivity as contact with surface residues. This is sup-ported by the feeding patterns documented in thebehavior observations in the Þeld-based bioassays.

Discussion

The objective of this research was to identify themodes of insecticidal activity critical to each com-poundÕs performance and to describe the spatial andtemporal dimension of these control mechanisms. Todo this we used a spectrum of evaluations, including

Fig. 2. Residue proÞles of Þve treatment chemicals onapple. Residues measured in micrograms per gram of activeingredient per leaf and fruit tissue taken at 4 h, 7 d, and 14 dpostÞeld application.


laboratory topical bioassays to determine acute con-tact activity and lethal time; Þeld-based residual bio-assays to identify lethal and sublethal modes of activitycontributing to adult control and fruit protection; be-havioral observations to identify temporal occur-rences of mortality, repellency, oviposition deter-rence, and antifeedant effects; and residue proÞleanalysis to understand the contribution of fruit andleaf surface and internal residues to these lethal andsublethal modes of insecticidal activity. No one formof evaluation alone can provide sufÞcient informationto fully describe the natural interactions between theplant, insect and chemical. Rather, each research com-ponent reveals a critical and complementary piece ofthe puzzle. We are convinced that this PIC model ofinvestigation is noteworthy and can be applied toother pests and crop systems. For simplicity, we canrefer to this model as the PIC triad, representing theimportance of measuring the three-way interactionbetween the plant, insect and chemistry.

The use of residue proÞle analysis in this initialresearch experience was limited in part by our de-pendence on composite residue samples. Future stud-ies can be strengthened with fully replicated residuesampling and further measurement of plant tissue at-tributes.

In the pursuit of crop protection, control of a pestcan be achieved through more than just lethal means.The basis for pest management with OPs such as azin-phosmethyl is reliance on their fast-acting contacttoxicity. In this study, we see that for many new in-secticides, other modes of activity are important con-tributors to the overall crop protection seen in theÞeld. In addition, these sublethal modes of activitymay work in concert with lethal modes based on atemporal sequence of activity. Residue proÞle analysisof the neonicotinoid insecticides thiamethoxam, imi-dacloprid, and thiacloprid reveals a dynamic interac-tion with fruit and leaf tissues that serves to regulatethe mode of activity at hand. Strong lethal activityoccurs as longas sufÞcient residuesarepresenton fruitand foliage surface. As these surface residues degrade,a range of sublethal behavioral effects, such as ovipo-sition deterrence, repellency, and antifeedant activityare generated in the presence of the remaining resi-

dues. The temporal sequence of lethal and sublethalmodes in this case is sufÞcient to provide the cropprotection demonstrated in Þeld efÞcacy trials. Otherstudies (Nauen et al. 1998, Isaacs et al. 1999) also havedocumented sublethal behavioral effects after expo-sure to or detection of internal plant residues. Twonicotinic receptor subtypes (desensitizing and non-desensitizing) have been identiÞed in cockroaches(Salgado and Saar 2004), one subtype causing acutelethal symptoms in the presence of high concentra-tions of imidacloprid (nAChN-type), and the othersubtype causing subacute symptoms under very lowconcentrations (nAChD-type). In addition to beingresponsible for antifeedant activity reported by Nauenand Salgado, nAChD receptor inhibition may also beresponsible for the oviposition deterrence observed inthis study.

Indoxacarb represents an example of the impor-tance of further characterizing the nature of lethalactivity to understand a compoundÕs performance.Although Þeld-based bioassays showed indoxacarb, ofall the tested OP-alternative compounds, to be themost similar to azinphosmethyl in terms of lethal ac-tivity, behavior observations and topical bioassayswere needed to properly describe the nature of itsactivity. The very high LD50 and LT50 values seen inthe topical bioassays indicate that the lethal activity ofindoxacarb is very slow compared with azinphos-methyl and that dermal contact alone is not an optimalmeans of delivering the poison. The plum curculiofeeding documented in the Þeld-based bioassays andbehavioral studies supports what others have pub-lished about ingestion being an important mode ofentry for the bioactivation of formulated DPX-JW062to its decarbomethoxylated metabolite (DCJW)(Wing et al. 1997, Tillman et al. 2001). The correlativerelationship between internal residues and mortality isan additional indicator that ingestion is important forenhanced lethal activity of indoxacarb. We cannotassume, however, that surface residues do not con-tribute to the toxicity of ingestion-activated com-pounds such as indoxacarb. Plum curculios have tochew through surface tissue residues to reach internalplant tissues, and dislogeable surface residues also canbe ingested through grooming behavior. This is the

Table 5. Regression R-Squared correlations between mortality (arcsine-transformed) and fruit and leaf residue for southern strainplum curculio

Treatment Surface leaf Surface fruit Surface total Interior leaf Interior fruit Interior total

Azinphosmethyl 0.8379** 0.4732* 0.6884** 0.6613* 0.4969* 0.7160**Imidacloprid 0.6347* 0.6440* 0.6353* 0.0574 0.0160 0.0483Thiacloprid 0.7550** 0.7431** 0.8955** Ña 0.4372* 0.4372*Thiamethoxam 0.7353** 0.7264** 0.7240** 0.7348** 0.5263* 0.7208**Indoxacarb 0.1083 0.1894 0.1499 0.1915 0.2347 0.1993Indoxacarbb 0.3479 0.5328* 0.4473* 0.5372* 0.6159* 0.5526*

Data are derived from Þeld-based bioassays and three parallel postapplication composite residue samples. All regression models have 1 df(model). Model used is mortality � residue. P values shown are for the overall model ANOVAs. *, model signiÞcant at � � 0.05; **, modelsigniÞcant at � � 0.001.a Analysis not possible owing to lack of variation in residue proÞle.bCorrelations for indoxacarb with suspect replication (4-h residue, no mortality) removed.


best explanation of the differences in mortality seen inthe Þeld-based bioassays compared with laboratory-based topical bioassays. The general term “contactactivity” is typically associated with the mortality re-sulting from insects being in contact with surface res-idues. This speciÞc mode of exposure may be furtherreÞned as “transcuticular activity” to more accuratelydescribe movement of the insecticide through theinsect cuticle to internal target sites. This is distinct,however, from inadvertent ingestion through insectgrooming behavior, which also may result in lethalactivity.

Identifying the modes of activity that contribute toinsecticide performance allows us to optimize the tim-ing of Þeld sprays for each compound based on itscritical performance characteristics. Azinphosmethylis the least sensitive to application timing because ofits contact toxicity and long-lasting surface residues.That is why petal fall sprays of this compound havebeen largely effective for control of the plum curculioin apples. Our data suggest that the optimal timing forneonicotinoids will be different from that of organo-phosphates. To capture both the short-lived lethalactivity and the longer lived oviposition deterrence ofthese compounds, it will important to cover as muchfruit surface area as possible. This can be best achievedby waiting several days after petal fall for fruit to beginsizing. Careful Þeld scouting of plum curculio for theinitiation of oviposition is also important so that fruitinjury does not occur before the neonicotinoids areapplied. To optimize the performance of indoxacarb,the application timing should be sequenced with plumcurculio feeding behavior. Because plum curculioadult feeding on plant tissue starts before the bloomperiod and diminishes as oviposition behavior ensues(unpublished data), traditional petal fall timing isprobably the best timing to maximize beetle presenceand feeding behavior.

The critical performance characteristics of a com-pound should include knowledge of its modes of ac-tivity, life stage activity for the target pest, mode ofexposure, and its residue proÞle on the respectivecrop. Compiling these characteristics for each insec-ticide, pest and crop will serve as a basis for conÞdentpest management decisions that optimize control,conserve beneÞcial species, and minimize costs. It alsocan serve to predict chemistry combinations with in-compatible modes of activity, such as neonicotinoids(anti-feedant activity) and indoxacarb (ingestion ac-tivity) or suggest as-yet unidentiÞed synergistic com-binations.

The introduction of several new classes of insecti-cides in the last decade has the potential for advancingintegrated pest management within fruit productionsystems. But there is increased risk for growers if thesenew tools are promoted as “OP replacements” that canbe reßexively switched into their spray programs with-out consideration of the diversity of performancecharacteristics among them. The burden is on theresearch community to thoroughly compile and ex-tend the knowledge in a meaningful manner for thelong-term sustainability of U.S. fruit production.

Acknowledgments

We thank Ryan VanderPoppen, Kevin Schoenborn, JasonSeward, and the farm staff of the Trevor Nichols ResearchComplex for technical assistance in carrying out this re-search. We also acknowledge the contribution of test sub-stances from the following agrichemical companies: BayerCropScience, Syngenta Crop Protection, and DuPont CropProtection. We thank the Michigan Apple Research Com-mittee for funding that helped to offset the costs of conduct-ing this research.

References Cited

Abbott, W. S. 1925. A model for computing the effective-ness of an insecticide. J. Econ. Entomol. 18: 265Ð267.

Borchert, D. M., J. F.Walgenbach, G. G. Kennedy, and J.W.Long. 2004. Toxicity and residual activity of methoxy-fenozide and tebufenozide to codling moth (Lepidop-tera: Tortricidae) and oriental fruit moth (Lepidoptera:Tortricidae). J. Econ. Entomol. 97: 1342Ð1352.

Chapman, P. J. 1938. The plum curculio as an apple pest.N.Y. State Agric. Exp. Stn. Bull. 684.

Hu, X. P., and R. J. Prokopy. 1998. Lethal and sublethaleffects of imidacloprid on apple maggot ßy, Rhagoletispomonella Walsh (Diptera: Tephritidae). J. Appl. Ento-mol. 122: 37Ð42.

Isaacs, R., M. Cahill, and D. N. Byrne. 1999. Host plantevaluation behavior ofBemisia tabaci and its modiÞcationby external or internal uptake of imidacloprid. Physiol.Entomol. 24: 1Ð8.

Kerchove, G. 2005. A statistical handbook for agricultureÞeld trials specialists. Gylling Data Management Inc.,Brookings, SD.

Kunkel, B., D. Held, and D. Potter. 2001. Lethal and sub-lethal effects of bendiocarb, halofenocide, and imidaclo-prid onHarpalus pennsylvanicus(Coleoptera: Carabidae)following different modes of exposure in turfgrass. J.Econ. Entomol. 94: 62Ð67.

Levine, E., and F. R. Hall. 1977. Effect of feeding and ovi-position by the plum curculio on apple and plum fruitabscission. J. Econ. Entomol. 70: 603Ð607.

Liburd, O. E., E. M. Finn, K. L. Pettit, and J. C. Wise. 2003.Response of blueberry maggot ßy (Diptera: Tephritidae)to imidacloprid-treated spheres and selected insecticides.Can. Entomol. 135: 427Ð438.

MacDonald, L., and T. Meyer. 1998. Determination of imi-dacloprid and triadimefon in white pine by gas chroma-tography/mass spectrometry. J. Agric. Food Chem. 46:3133Ð3138.

Nauen, R. 1995. Behavior modifying effects of low systemicconcentrations of imidacloprid on Myzus persicae withspecial reference to an antifeeding response. Pestic. Sci.44: 145Ð153.

Nauen, R., B. Koob, andA. Elbert. 1998. Antifeedant effectsof sublethal dosages of imidacloprid on Bemisia tabaci.Entomol. Exp. Appl. 88: 287Ð293.

Prokopy, R. J., G. Galarza, and P. L. Phelan. 1993. Monitor-ing plum curculio in orchards: new hope for a moreeffective method. Proc. New England Fruit Meet. 91:70Ð76.

Racette, G., G. Chouinard, C. Vincent, and S. B. Hill. 1992.Ecology and management of the plum curculio, Conotra-chelus nenuphar (Coleoptera: Curculionidae), in appleorchards. Phytoprotection 73: 85Ð100.

Salgado, V., and R. Saar. 2004. Desensitizing and non-de-sensitizing subtypes of alpha-bungarotoxin-sensitive nic-


otinic acetylcholine receptors in cockroach neurons.J. Insect Physiol. 50: 867Ð879.

SAS Institute. 2002. PROC userÕs manual, version 7th ed.SAS Institute, Cary, NC.

Sokal, R. R., and F. J Rohlf. 1994. Biometry: the principlesand practice of statistics in biological research, 3rd ed.Freeman, New York

Smith, E. H. 1957. A method for rearing the plum curculiounder laboratory conditions including some biologicalobservations. J. Econ. Entomol. 50: 187Ð190.

Stedman, J. M. 1904. The “sting” in the appleÐthe work ofthe plum curculio in the apple. Univ. Mo. Agric. Exp. Sta.Bull. 64: 1Ð24.

Tedders,W.L., andB.W.Wood. 1994. A new technique formonitoring pecan weevil emergence (Coleoptera: Cur-culionidae). J. Entomol. Sci. 29: 18Ð30.

Tillman P. G., G. Hammes,M. Sacher,M. Connair, E. Brady,and K. Wing. 2001. Toxicity of a formulation of theinsecticide indoxacarb to the tarnished plant bug, Lygus

lineolaris (Hemiptera: Miridae), and the big-eyed bug,Geocoris punctipes (Hemiptera: Lygaeidae). Pest Manag.Sci. 58: 92Ð100.

Wamhoff, H., and V. Schneider. 1999. Photodegradation ofImidacloprid. J. Agric. Food Chem. 47: 1730Ð1734.

Wing K.D., M. Schnee,M. Sacher, andM. Connair. 1997. Anovel oxadiazine insecticide is bioactivated in lepidop-teran larvae. Arch. Insect Biochem. Physiol. 37: 91Ð103.

Wise, J.C., andL.Gut. 2000. Control of plum curculio, 1999.Arthropod Manag. Tests 25 (A26). (http: //www.entsoc.org/Protected/�MT/AMT25/A24.htm).

Yonce, C. E., D. L. Horton, and W. R. Okie. 1995. Springmigration, reproductivebehavior,monitoringproceduresand host preference of plum curculio (Coleoptera: Cur-culionidae) on Prunus species in central Georgia. J. En-tomol. Sci. 30: 82Ð92.

Received 17 March 2006; accepted 7 August 2006.


Documents

H E Use of Residue Proﬁle Analysis to Identify Modes of ...Gut 2000). Although such Þeld trials are encouraging in terms of showing basic fruit protection over pre-scribed intervals,