BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors nonprofitpublishers academic institutions research libraries and research funders in the common goal of maximizing access tocritical research

Role of Syrphid Larvae and Other Predators in SuppressingAphid Infestations in Organic Lettuce on Californiarsquos CentralCoastAuthor(s) Hugh A Smith William E Chaney Tiffany A BensenSource Journal of Economic Entomology 101(5)1526-1532 2008Published By Entomological Society of AmericaDOI httpdxdoiorg1016030022-0493(2008)101[1526ROSLAO]20CO2URL httpwwwbiooneorgdoifull1016030022-0493282008291015B15263AROSLAO5D20CO3B2

BioOne (wwwbiooneorg) is a nonprofit online aggregation of core research in thebiological ecological and environmental sciences BioOne provides a sustainable onlineplatform for over 170 journals and books published by nonprofit societies associationsmuseums institutions and presses

Your use of this PDF the BioOne Web site and all posted and associated contentindicates your acceptance of BioOnersquos Terms of Use available at wwwbiooneorgpageterms_of_use

Usage of BioOne content is strictly limited to personal educational and non-commercialuse Commercial inquiries or rights and permissions requests should be directed to theindividual publisher as copyright holder

BIOLOGICAL AND MICROBIAL CONTROL

Role of Syrphid Larvae and Other Predators in Suppressing AphidInfestations in Organic Lettuce on Californiarsquos Central Coast

HUGH A SMITH12 WILLIAM E CHANEY3 AND TIFFANY A BENSEN4

J Econ Entomol 101(5) 1526ETH1532 (2008)

ABSTRACT Organic lettuce Lactuca sativa L growers on the Central Coast of California rely onconservation biological control to manage Nasonovia ribisnigriMosley (Hemiptera Aphididae) andother aphid pests of lettuce In 2006 we carried out THORNve replicated THORNeld trials to determine theimportance of syrphid larvae in the suppression of N ribisnigri and other aphids infesting organicromaine lettuce We used Entrust a spinosad-based insecticide approved for use on organic farms tosuppress syrphid larvae in aphid-infested romaine Romaine treated with Entrust was unmarketableat harvest because of aphid infestation whereas insecticide-free romaine was marketable Syrphidlarvae composed 85 or more of total predators in most trials and they were the only predatorsconsistently recovered from romaine that was infested with aphids early and largely aphid-free byharvest The species mix of nonsyrphid predators varied from site to site Applications of Entrustsuppressed nonsyrphid predators in two trials and so was an imperfect tool for selectively suppressingsyrphid larvae The relative importance of syrphid larvae and other predators in the conservationbiological control of aphids in organic romaine is discussed We conclude that syrphid larvae areprimarily responsible for the suppression of aphids in organic romaine on CaliforniaOtildes Central Coast

KEY WORDS conservation biocontrol organic agriculture insectary crops aphid predators Syr-phidae

In 2005 200000 acres of lettuce (head and leaf)were grown in California which produces roughlythree quarters of the lettuce consumed in the UnitedStates (NASS 2006) More than 14000 acres of organicleaf lettuces were produced in California in 2005(Klonsky and Richter 2007) At least 70 of Califor-niaOtildes lettuce both conventional and organic is grownin the stateOtildes Central Coast region (California FarmBureau 2006 Klonsky and Richter 2007)Nasonovia ribisnigri (Mosley) (Hemiptera Aphidi-

dae) became established in California in 1998 and ithas become the most important insect pest of lettuceon the Central Coast (Chaney 1999) N ribisnigriinfests the inner leaves of the lettuce head making itunmarketable (MacKenzie and Vernon 1988 Liu2004) In addition to N ribisnigri the foxglove aphidAulocorthum solani (Kaltebach) green peach aphidMyzus persicae (Sulzer) and potato aphid Macrosi-phum euphorbiae (Thomas) are pests of Californialettuce Conventional growers suppress populations ofN ribisnigri and other aphids with a range of insecti-

cides including organophosphate carbamate andneonicotinoid materials

Fieldwork in 2005 revealed that at least 13 speciesof syrphid larvae are involved in suppressing aphidinfestations to below economically damaging levels inorganically grown lettuce on CaliforniaOtildes CentralCoast (Smith and Chaney 2007) Organic growerstypically interplant szligowering insectary plants withlettuce to provide szligoral resources to syrphid adultsand enhance their activity Other predators are foundin organically grown romaine in this region but theirimportance relative to syrphid larvae in suppressingaphids has not been demonstrated previously (Colfer2004 Smith and Chaney 2007) Parasitism ofN ribisni-gri is rarely observed because the aphid colonizesinner portions of the plant that are largely inaccessibleto parasitoids Green peach aphid and potato aphidtend to infest theouter leavesof the romainehead andthey are parasitized Entomogenous fungi have beenobserved to impact infestations of N ribisnigri duringthe early spring rains but rarely during peak lettuceproduction (MayETHNovember) (Smith and Chaney2007)

Our objective in this study was to evaluate theimportance of syrphid predation in suppressing pop-ulations ofN ribisnigri and other aphids in organicallygrown romaine on CaliforniaOtildes Central Coast In ad-dition we were interested in the importance of otherpredators relative to syrphid larvae in suppressingaphids We used Entrust a spinosad-based insecticide

1 Corresponding author University of California Cooperative Ex-tension 624 West Foster Rd Santa Maria CA 93455

2 Current address Connecticut Agricultural Experiment Station153 Cook Hill Road Windsor CT 06095 (HughSmithpostatectus)

3 University of California Cooperative Extension (emeritus) 1432Abbott St Salinas CA 93901

4 Department of Biology PO Box 1848 University of MississippiUniversity MS 38677

0022-0493081526ETH1532$04000 2008 Entomological Society of America

approved for use in organic lettuce to suppress syr-phid larvae We compared aphid densities in romainewhere syrphids were suppressed with aphid densitiesin insecticide-free romaine Pilot studies in previousyears indicated that aphid densities would not be af-fected by Entrust at the rate used We did not knowwhat impact weekly applications of Entrust wouldhave on the complex of nonsyrphid predators found inorganically grown romaine although some studieshave indicated that spinosad has little suppressive ef-fect on predators such as ladybird beetles (Coccine-lidae) lacewings (Chrysopidae and Hemerobiidae)and minute pirate bugs (Orius spp) (Eizen et al 1998Studebaker and Kring 2003 Miles 2006 Contreras etal 2006 Arthurs et al 2007)

Although the focus of this study is the evaluation ofconservation biological control of aphids in organi-cally grown romaine conservation biological controlmay eventually be applicable in conventional produc-tion There is currently little role for natural enemiesin the suppression of N ribisnigri in conventionallettuce because conventional growers rely on insec-ticide regimes that are not compatible with biologicalcontrol Industry standards for conventional producealso make the risk of relying on natural enemies un-acceptably high for conventional growers Howeverthe registrations of many organophosphate and car-bamate insecticides are currently under review by theUS Environmental Protection Agency Should theinsecticides that conventional lettuce growers cur-rently rely on for aphid suppression become lessavailable in the future information on conservationbiological control of aphids may be useful to con-ventional growers as well as organic growers There-fore clariTHORNcation of the mechanism by which aphidsare suppressed in organic lettuce may enhance fu-ture prospects for conservation biological control inconventional production

Materials and Methods

We suppressed syrphid larvae with the spinosad-based insecticide Entrust to evaluate how aphid pop-ulations responded to release from syrphid predationWe compared densities of aphids syrphid eggs andlarvae and other natural enemies of aphids in plots oforganic romaine lettuce treated with Entrust to den-sities in untreated plots at THORNve sites in and near theSalinas Valley in 2006 Separate trials were conductedin THORNelds on certiTHORNed organic farms in Hollister (AprilETHMay) Chualar (AugustETHSeptember) and Watsonville(SeptemberETHOctober) and in research plots at theHartnell College East Campus (JuneETHJuly) and theSpreckels Industrial Park (JulyETHAugust)

Insect densities on romaine receiving no insecticideapplications were compared with insect densities onromaine receiving the equivalent of 0182 liters En-trust per hectare (250 ozacre) each week appliedwith a pressurized CO2 backpack sprayer Treatmentswere applied in 762-m (25-foot) sections of crop rowreplicated four times and arranged in randomizedcomplete blocks Trials on organic farms were con-

THORNned to one section of the THORNeld (two beds wide and100 feet long) At the Hartnell and Spreckels sites thetrials were carried out in plots of the same dimensionEntrust treatments were initiated 1 wk after trans-planting or thinning of romaine We began samplingthe next week Plots were sampled once a week untilharvest

On the three organic farms we evaluated naturalinfestations of N ribisnigri and other aphids attackingromaine For the Hartnell and Spreckels trials we trans-planted lettuce seedlings infested with N ribisnigriTrays of lettuce transplants were conTHORNned for 3 d withina laboratory colony of N ribisnigri to establish an infes-tation before transplanting at these two THORNeld trials

Three whole romaine plants were sampled fromeach plot each week for a total of 12 plants per treat-ment per week Plants were placed in plastic bags andtaken to the University of California Cooperative Ex-tension laboratory in Salinas for processing An 189-liter (5-gal) plastic water bottle was modiTHORNed to func-tion as a collection sieve for insects on the romaineThe lower half of the bottle was cut away and thecentral portion of the screw-on cap was replaced withorgandy cloth by using a glue gun The half-bottle wasplaced in a large sink with the cap end down Plantswere washed over the cut-away end of the bottle sothat the water carried all insects into the screw-on capwhere material collected on the organdy Plants werewashed using a hand-held showerhead-type faucetThe cap was removed after the plant had been thor-oughly washed and its contents were examined undera microscope The number and species of aphid wererecorded for each plant as was the number of syrphideggs and larvae the number of parasitized aphids andthe number and type of other predators

Syrphid larvae were collected from samples andplaced individually in a petri dish (100 by 25 cmThermo Fisher ScientiTHORNc Waltham MA) for rearingThe larvae were kept in a growth chamber (modelMB60-B Percival ScientiTHORNc Perry IA) with a photo-period of 168 (LD) h at 20ETH25 C and 60ETH75 RH andthey were provided with N ribisnigri pea aphidandor foxglove aphid on pieces of lettuce leaf or fababeanVicia fabaL leaf every 3 d until pupation Uponreaching the adult stage syrphids were allowed 1 d toexpel abdominalszliguid then theywere frozen and laterpinned and placed in a reference collection The spe-cies of each individual was recorded

To test for differences in whole plant densities ofaphids syrphid larvae nonsyrphid predators and syr-phid eggs between the two treatments over time andamong trials we analyzed together the THORNnal 5 wk ofdata from each trial by using a repeated measuresanalysis of variance (ANOVA) model Fixed factors inthe model were trial and treatment (between subject)and sampling week (within subject) The random ef-fects of block (trial) and treatment block (trial)were included when associated covariance parame-ters were estimated to be greater than zero SigniTHORNcantinteractions between factors were followed with testsof simple effects of treatment within trial or weekwithin trial

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1527

Aphid parasitism was determined by counts ofaphid mummies We tested for differences in parasit-ism between treatments and over time separately foreach trial with a repeated measures ANOVA modelTreatment (between-subject) and sampling week(within-subject) were THORNxed factors in the model andthe random effect of block was included when theassociated covariance parameter was estimated to begreater than zero SigniTHORNcant interactions were fol-lowed with tests of simple effects of treatment withinweek

To examine the numerical relationships betweenaphid and syrphid larva densities and between aphidand nonsyrphid predator densities across all trials weapplied a multiple linear regression model to the fulldata set This model included trial and week as THORNxedcategorical variables syrphid larvae and nonsyrphidpredators as THORNxed continuous variables and interac-tions between trial and each continuous variable

All analyses were performed using the MIXEDprocedure of SAS (SAS Institute 2002ETH2003) whichestimates parameters using restricted maximum like-lihood A THORNrst-order autoregressive covariance modelwas used to incorporate correlations in the data col-lected over time Denominator degrees of freedomwere adjusted using the KenwardETHRogers calculationWhen necessary to meet the assumption of normalitydata were transformed using a natural log transforma-tion although untransformed least squares means arepresented to facilitate interpretation Statistical sig-niTHORNcance for all tests was deTHORNned as 005

Results

Aphid Densities N ribisnigri was the predominantaphid species found in each trial although foxgloveaphid potato aphid and green peach aphid also con-tributed to infestations Whole plant densities ofaphids were higher in plots treated with Entrust thanin untreated plots in each trial usually beginning onthe third or fourth sampling week of a trial (trial treatment week interaction Table 1) The highestaverage whole plant aphid infestation levels in ro-maine treated with Entrust were as much as 36 timeshigher than peak infestation levels in untreated romainein the same trial (Table 2) At-harvest aphid infestationlevels for romaine treated with Entrust ranged from2802 922 (mean SEM) per plant in the Hollistertrial to 14003 922 in the Spreckels trial Romainetreated with Entrust was unmarketable because of aphidinfestation In contrast average aphid densities per plantat harvest in untreated romaine ranged from 219 922in the Hartnell trial to 948 922 at the Spreckels siteUntreated romaine was marketableSyrphid Densities Syrphid larvae composed be-

tween 85 and 96 of predators collected from un-treated romaine in the Chualar Hartnell Hollisterand Watsonville trials and 63 of the predators col-lected from untreated romaine at Spreckels (Table 3)Syrphid larvae were the only predator group recov-ered from untreated romaine at all THORNve sites Usuallyby the third week and always by the THORNnal (harvest)week of sampling for each trial whole plant densitiesof syrphids were signiTHORNcantly higher in untreated plots

Table 1 Results from ANOVA comparing whole plant densities of aphids syrphid larvae syrphid eggs and nonsyrphid predatorsbetween untreated plots and plots treated with Entrust and across trials and time

Factors and sources of variationdf

F statistic P valueNumerator Denominator

Aphid densityTrial 4 301 3700 00001Treatment 1 301 24844 00001Trial treatment 4 301 444 00062Sampling wk 4 218 3384 00001Treatment sampling wk 4 218 5910 00001Trial treatment sampling wk 32 218 1018 00001

Syrphid densityTrial 4 15 807 00011Treatment 1 15 28336 00001Trial treatment 4 15 646 00031Sampling wk 4 222 8725 00001Treatment sampling wk 4 222 3751 00001Trial treatment sampling wk 32 222 1098 00001

Syrphid egg densityTrial 4 15 3153 00001Treatment 1 534 1718 00001Trial treatment 4 534 168 01539Sampling wk 4 534 4879 00001Treatment sampling wk 4 534 604 00001Trial treatment sampling wk 32 534 1203 00001

Nonsyrphid predator densityTrial 4 136 323 00144Treatment 1 136 869 00038Trial treatment 4 136 851 00001Sampling wk 4 281 3284 00001Treatment sampling wk 4 281 621 00001Trial treatment sampling wk 32 283 223 00003

1528 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

compared with Entrust plots (trial treatment week interaction Table 1) The highest average wholeplant syrphid densities in untreated romaine rangedfrom 275 058 (at Spreckels during the last week)to 908 058 (at Hartnell in the fourth week) (Table4) The highest average whole plant syrphid densitiesin romaine treated with Entrust ranged from 017 058 (at Spreckels in the fourth week) to 284 058(at Hartnell also in the fourth week)Syrphid Egg Densities Average whole plant densi-

ties of syrphid eggs differed statistically between treat-ments in three of the THORNve trials but only during certainweeks (trial treatment week interaction Table1) Average whole plant densities of syrphid eggs weresigniTHORNcantly higher in romaine treated with Entrustthan in untreated romaine during weeks 4 and 5 of theWatsonville trial (week 4 Entrust 870 074 un-

treated 282 074 week 5 Entrust 592 074untreated 053 074) and during the last samplingweek at the Hartnell trial (Entrust 184 074untreated 042 074) Syrphid egg densities werehigher in untreated romaine during the third week atthe Spreckels trial (Entrust 008 074 untreated 125 074) Whole plant densities of syrphid eggsaveraged 246 026 at the Chualar trial and 297 026 at the Hollister trialSyrphid Species Eight syrphid species and one syr-

phid parasitoid species (Diplazon sp Ichneu-monidae) were reared from larvae collected duringthe course of these trials Each of these species hadbeen recovered during a survey of syrphids in organicromaine in 2005 (Smith and Chaney 2007) with theexception of Sphaerophoria contigua Maquart Spha-erophoria sulfuripes (Thomson) Allograpta obliqua

Table 2 Simple effect test results and untransformed least squares means SEM for whole plant aphid densities between treatmentswithin each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 1679 922 5001 922 11613 922 4355 922 5285 922Untreated 1593 922 1653 922 2235 922 242 922 247 922F P value 056 0456 1683 00001 3382 00001 5961 00001 6012 00001

HollisterEntrust 2367 922 6301 922 6529 922 6378 922 2802 922Untreated 1912 922 6977 922 6097 922 072 922 605 924F P value 003 0859 002 0877 000 0992 9515 00001 1227 0001

HartnellEntrust 441 922 1038 922 2162 922 2864 922 7985 922Untreated 148 922 543 922 1456 922 136 922 219 922F P value 351 0061 301 0083 883 0003 4955 00001 10025 00001

WatsonvilleEntrust 4271 922 5135 922 6425 922 16119 922 6424 922Untreated 4042 922 3334 922 3298 922 871 922 570 922F P value 003 0871 315 0077 399 0046 6715 00001 5814 00001

SpreckelsEntrust 580 922 954 922 3951 922 9684 922 14003 922Untreated 243 922 1663 922 5074 922 2389 922 948 922F P value 243 0120 003 0853 003 0867 1997 00001 6559 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

Table 3 Abundance of all predators and percentages of predator groups in each treatment by trial

TrialTotalno

total no

Syrphidlarvae

Dwarf spiderMinute pirate

bugLadybird beetle Lacewing Rove beetle Bigeyed bug

ChualarUntreated 258 85 7 3 Ntildea Ntilde 4 NtildeEntrust 16 81 13 6 Ntilde Ntilde Ntilde Ntilde

HollisterUntreated 189 88 Ntilde Ntilde 10 2 Ntilde NtildeEntrust 77 47 1 Ntilde 34 18 Ntilde Ntilde

HartnellUntreated 233 96 3 Ntilde 1 1 Ntilde NtildeEntrust 65 80 8 Ntilde 8 4 Ntilde Ntilde

WatsonvilleUntreated 229 91 2 1 Ntilde 1 6 NtildeEntrust 50 40 10 8 4 6 32 Ntilde

SpreckelsUntreated 109 63 2 30 4 1 Ntilde NtildeEntrust 17 18 Ntilde 12 53 12 Ntilde 6

aNtilde not collected

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1529

(Say) Toxomerus marginatus (Say) and Allograptaexotica Wiedemann were the most common speciesreared Syrphus opinatorOsten SackenPlatycheirus steg-nus (Say) Eupeodes americanusWiedemann and Spha-erophoria continua were recovered in low numbersNonsyrphid Predators Whole-plant densities of

nonsyrphid predators generally were low in all trialsAverage whole plant densities of nonsyrphid preda-tors ranged from zero in multiple weeks to 174 021in romaine treated with Entrust during the THORNnal weekof sampling at the Hollister site Entrust affected non-syrphid predators differently depending on trial andsampling week (trial treatment week interaction

Table 1) At the Chualar and Spreckels trials nonsyr-phid predator densities were signiTHORNcantly higher inthe untreated lettuce compared with romaine treatedwith Entrust for the last 3 and 2 weeks of samplingrespectively (Table 5) In contrast at the Hollister andWatsonville trials nonsyrphid predators were signif-icantly higher in romaine treated with Entrust than inuntreated romaine during the week before harvestNonsyrphid predators did not differ statistically be-tween treatments at the Hartnell trial (Table 5)

Nonsyrphid predators collected from organic ro-maine included dwarf spiders (Linyphiidae) minutepirate bugs (Orius sp) ladybird beetles (Coccineli-

Table 4 Simple effect test results and untransformed least squares means SEM for whole plant syrphid larva densities betweentreatments within each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 000 058 050 058 025 058 033 058 000 058Untreated 025 058 329 058 896 058 350 058 232 058F P value 044 0509 2942 00001 12734 00001 3569 00001 2408 00001

HollisterEntrust 000 058 000 058 009 058 073 058 219 058Untreated 000 058 000 058 076 058 649 058 666 058F P value 000 1000 000 0996 240 00001 6562 00001 3758 00001

HartnellEntrust 009 058 000 058 125 060 284 058 025 058Untreated 000 058 066 058 286 058 908 058 598 058F P value 005 0825 285 0094 868 0004 4431 00001 8076 00001

WatsonvilleEntrust 017 058 042 058 025 058 041 058 042 058Untreated 076 058 281 058 530 058 546 058 334 058F P value 211 0149 1922 00001 4336 00001 6067 00001 3255 00001

SpreckelsEntrust 000 058 000 058 000 058 017 058 008 058Untreated 008 058 026 058 100 058 168 058 275 058F P value 005 0821 044 0510 433 0040 113 0001 2904 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

Table 5 Simple effect test results and untransformed least squares means SEM for whole plant nonsyrphid predator densitiesbetween treatments within each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 007 021 000 021 000 021 001 021 016 021Untreated 010 021 016 021 063 021 108 021 116 021F P value 002 0893 076 0384 1001 0002 2123 00001 745 0007

HollisterEntrust 000 021 000 021 034 021 134 021 174 021Untreated 000 021 001 021 044 021 089 021 051 021F P value 000 0994 001 0910 000 0991 042 0516 1744 00001

HartnellEntrust 000 021 009 021 000 021 007 021 093 021Untreated 000 021 000 021 000 021 041 021 045 021F P value 000 0952 027 0610 000 0945 166 0199 316 0077

WatsonvilleEntrust 024 021 000 021 000 021 025 021 125 021Untreated 022 021 002 021 016 021 065 021 036 021F P value 057 0450 001 0920 082 0365 147 0226 622 0013

SpreckelsEntrust 000 021 016 021 001 021 000 021 000 021Untreated 000 021 000 021 040 021 103 021 134 021F P value 000 0974 091 0340 252 0114 1592 00001 2281 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

1530 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

dae) brown lacewings (Hemerobiidae) rove beetles(Staphylinidae) and big-eyed bugs (Geocoris sp)The complex of nonsyrphid predators was different ateach site (Table 3)ParasitismAphid mummies were found only in the

Hartnell and Watsonville trials Average whole plantdensities of aphid mummies were higher in romainetreated with Entrust (143 022) than in untreatedromaine (008 022) at the Hartnell trial (F 1871df 1 6 P 0005) At the Watsonville trial averagewhole plant densities of aphid mummies were signif-icantly higher in romaine treated with Entrust only inweek 5 (Entrust 325 049 untreated 020 049 F 2113 df 1 6 P 0004)Aphid Density Relationships with Syrphid Larvaeand Nonsyrphid Predators Regression analysis re-vealed a signiTHORNcant negative relationship betweenaphid and syrphid densities in all trials but a signiTHORNcantnegativerelationshipbetweennonsyrphidpredatorsandaphiddensities in theChualar trialonly(Table6)Of thefourothertrialswefoundnorelationshipbetweenaphidand nonsyrphid predator densities

Discussion

On CaliforniaOtildes Central Coast several differenttypes of predators are found in organically grownromaine that is infested with N ribisnigri and otheraphids The data presented here and previously(Smith and Chaney 2007) indicate that syrphid larvaeare the most abundant predators often making up85 of the predators collected from infested ro-maine and that they can average up to nine larvae perromaine head during the crucial weeks immediatelybefore the harvest of the crop In addition syrphidlarvae are the only predator group consistently foundin all infested romaine THORNelds Other predators are lessabundant and they are recovered in some THORNelds butnot others Regression analysis presented here sug-gests that increases in syrphid larvae help explainreductions in aphid numbers and that there is no

relationship between nonsyrphid predators and aphiddensities at most sites

We present data that demonstrates that aphid den-sities were signiTHORNcantly higher in plots where syrphidswere suppressed Romaine in which syrphids weresuppressed was unmarketable because of aphid infes-tation whereas untreated romaine was marketableOur results suggest that syrphids play a major role inthe conservation biological control ofN ribisnigri andother aphids in organically grown romaine on theCentral Coast of California

We found that although Entrust does not affectoviposition by syrphid females it does suppress non-syrphid predators in some instances Because Entrustapparently reduced the densities of nonsyrphid pred-ators as well as syrphid larvae at the Chualar andSpreckels trials it was an imperfect tool for demon-strating the effect of selectively removing syrphid lar-vae from organically grown romaine Given thatminute pirate bugs composed 30 of the predators atthe Spreckels trial in untreated romaine and only 12where Entrust was applied it is possible that suppres-sion of minute pirate bugs contributed to the releaseof aphids from predation in the Entrust treatmentDwarf spiders and minute pirate bugs in the Chualartrial also may have contributed to aphid suppressionDwarf spiders are important predators of aphids inwheat Triticum aestivum L (Harwood et al 2004Nienstedt and Poehling 2004) and minute pirate bugsare important predators of soybean aphid (Costama-gna and Landis 2007)

The Chualar and Spreckels trials illustrate instancesinwhichpredatorsother thansyrphidsmaycontributesigniTHORNcantly to the suppression of aphids in organicromaine However unlike the Chualar site whereaverage syrphid densities were among the highest ofall trials 2 wk before harvest syrphid densities werevery low at Spreckels The Spreckels trial also wasunusual in that the aphid infestation in the untreatedplots at harvest approached unmarketable levels Con-sidering the paucity of syrphid larvae the abundance

Table 6 Multiple linear regression analysis and parameter estimates SE for relationships between aphid (y) and syrphid larva (x)densities and between aphid (y) and nonsyrphid predator (z) densities

Factors and sources ofvariation

dfF statistic P value

Numerator Denominator

Trial 4 864 561 0001Syrphid Larvae 1 542 1792 00001Nonsyrphid predators 1 493 000 0982Syrphid larvae trial 4 567 142 0226Nonsyrphid predators trial 4 512 260 0035Sampling wk 4 467 1015 00001

SigniTHORNcance of slope parameter

Parameter estimates DF t value P value

Syrphid larvay 288(009) 0091(002)x 542 423 00001Nonsyrphid predator trialChualar y 298(021) 0302(011)z 505 278 0006Hollister y 330(021) 0080(009)z 486 090 0371Hartnell y 218(021) 0018(017)z 487 011 0914Watsonville y 349(022) 0018(012)z 510 015 0882Spreckels y 288(021) 0179(013)z 546 134 0181

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1531

of minute pirate bugs and the poor biological controlof aphids achieved at this site the Spreckels trial mayoffer an example of interguild predation interferingwith suppression of aphids by syrphids in organicromaine Interguild predation interferes with conser-vation biological control in some instances (Rosen-heim 1998) and further study is required to determinewhether the presence of certain nonsyrphid predatorsreduces the efTHORNcacy of syrphids in suppressing aphidsin organic romaine

It is worth noting that nonsyrphid predators werestatistically higher in romaine treated with Entrust inthe Hollister and Watsonville trials but that Entrust-treated romaine was still unmarketable in these trialsbecause of aphid infestation Numbers of aphid mum-mies also were higher in romaine treated with Entrustin two trials presumably because the densities ofaphids were much higher in romaine treated withEntrust

Given the abundance of syrphid larvae and thatsyrphid larvae are the only predator consistently re-covered from organically grown romaine we arguethat syrphid larvae are primarily responsible for sup-pressing N ribisnigri and other aphids to below eco-nomic levels

Acknowledgments

We thank Robert Bugg (University of California Sustain-able Agriculture Research and Development Program) andDiana Henderson Reisa Bigelow Franklin Dlott Cathy Carl-son Lourdes Rodriguez and Jason Hoeksema for assistancewith this research Sanford Eigenbrode and three anonymousreviewers helped improve the original manuscript We alsothank the farm managers who collaborated with on-farmtrials This research was supported in part by the UnitedStates Environmental Protection Agency Region 9

References Cited

Arthurs S P L A Lacey and E R Miliczky 2007 Eval-uation of codling moth granulosis virus and spinosad forcodling moth control and impact on non-target species inpear orchards Biol Control 41 99ETH109

California Farm Bureau 2006 California Farm Bureau(wwwcfbfcominfoagfacftscfm)

Chaney W E 1999 Lettuce aphid update MontereyCounty Crop Notes University of California CooperativeExtension Salinas CA

Colfer R 2004 Using habitat management to improvebiological control on commercial organic farms in Cal-

ifornia pp 55ETH62 In M S Hoddle [ed] Fourth Cal-ifornia Conference on Biological Control 13ETH15 July2004 Berkeley CA University of California PressBerkeley CA

Contreras J V Quinto V Abellan J Fernandez E Grava-los C Moros and P Bielza 2006 Effects of naturalinsecticides on Frankliniella occidentalis and Orius sppBull OILBSROP 29 331ETH336

Costamagna A C andDA Landis 2007 Quantifying pre-dation on soybean aphid through direct THORNeld observa-tions Biol Control 42 16ETH24

Eizen G W P J Elzen and E G King 1998 Laboratorytoxicity of insecticide residues to Orius insidiosus Geo-coris punctipes Hippodamia convergens and Chrysoperlacarneapp 1235ETH1238 InP Dugger and D Richter [eds]1998 Proceedings of the Beltwide Cotton Conference5ETH9 January 1998 San Diego CA

Harwood J D K D Sunderland and WOC Symondson2004 Prey selection by linyphiid spiders moleculartracking of the effects of alternative prey on rates of aphidconsumption in the THORNeld Mol Ecol 13 3549ETH3560

Klonsky K and K Richter 2007 Statistical review of Cal-ifornia Organic Agriculture 2000ETH2005 Agricultural Is-sues Center University of California Davis CA (aicuc-davisedu)

Liu Y-B 2004 Distribution and population developmentof Nasonovia ribisnigri in iceberg lettuce J Econ Ento-mol 97 883ETH890

MacKenzie J R and R S Vernon 1988 Sampling for thedistribution of lettuce aphid Nasonovia ribisnigri (Ho-moptera Aphididae) in THORNelds and within heads J En-tomol Soc Br Columbia 85 10ETH14

Miles M 2006 The effects of spinosad on beneTHORNcial insectsand mites used in integrated pest management systems ingreenhouses Bull OILSROP 29 53ETH59 [NASS] Na-tional Agricultural Statistics Service 2006 National Ag-ricultural Statistics Service (wwwnassusdagov)

Nienstedt K M and H-M Poehling 2004 Invertebratepredation of 15N-marked prey in semi-THORNeld wheat enclo-sures Entomol Exp Appl 112 191ETH200

Rosenheim J A 1998 High order predators and the regu-lation of insect herbivore populations Annu Rev Ento-mol 43 421ETH47

SAS Insitute 2002ndash2003 SAS version 91 SAS InstituteCary NC

Smith H A and W E Chaney 2007 A survey of syrphidpredators ofNasonovia ribisnigri in organic lettuce on theCentral Coast of California J Econ Entomol 100 39ETH48

Studebaker G E and T J Kring 2003 Effects of insecti-cides on Orius insidiosus measured by THORNeld greenhouseand Petri dish bioassays Fla Entomol 86 178ETH185

Received 21 August 2007 accepted 27 April 2008

1532 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

approved for use in organic lettuce to suppress syr-phid larvae We compared aphid densities in romainewhere syrphids were suppressed with aphid densitiesin insecticide-free romaine Pilot studies in previousyears indicated that aphid densities would not be af-fected by Entrust at the rate used We did not knowwhat impact weekly applications of Entrust wouldhave on the complex of nonsyrphid predators found inorganically grown romaine although some studieshave indicated that spinosad has little suppressive ef-fect on predators such as ladybird beetles (Coccine-lidae) lacewings (Chrysopidae and Hemerobiidae)and minute pirate bugs (Orius spp) (Eizen et al 1998Studebaker and Kring 2003 Miles 2006 Contreras etal 2006 Arthurs et al 2007)

Although the focus of this study is the evaluation ofconservation biological control of aphids in organi-cally grown romaine conservation biological controlmay eventually be applicable in conventional produc-tion There is currently little role for natural enemiesin the suppression of N ribisnigri in conventionallettuce because conventional growers rely on insec-ticide regimes that are not compatible with biologicalcontrol Industry standards for conventional producealso make the risk of relying on natural enemies un-acceptably high for conventional growers Howeverthe registrations of many organophosphate and car-bamate insecticides are currently under review by theUS Environmental Protection Agency Should theinsecticides that conventional lettuce growers cur-rently rely on for aphid suppression become lessavailable in the future information on conservationbiological control of aphids may be useful to con-ventional growers as well as organic growers There-fore clariTHORNcation of the mechanism by which aphidsare suppressed in organic lettuce may enhance fu-ture prospects for conservation biological control inconventional production

Materials and Methods

We suppressed syrphid larvae with the spinosad-based insecticide Entrust to evaluate how aphid pop-ulations responded to release from syrphid predationWe compared densities of aphids syrphid eggs andlarvae and other natural enemies of aphids in plots oforganic romaine lettuce treated with Entrust to den-sities in untreated plots at THORNve sites in and near theSalinas Valley in 2006 Separate trials were conductedin THORNelds on certiTHORNed organic farms in Hollister (AprilETHMay) Chualar (AugustETHSeptember) and Watsonville(SeptemberETHOctober) and in research plots at theHartnell College East Campus (JuneETHJuly) and theSpreckels Industrial Park (JulyETHAugust)

Insect densities on romaine receiving no insecticideapplications were compared with insect densities onromaine receiving the equivalent of 0182 liters En-trust per hectare (250 ozacre) each week appliedwith a pressurized CO2 backpack sprayer Treatmentswere applied in 762-m (25-foot) sections of crop rowreplicated four times and arranged in randomizedcomplete blocks Trials on organic farms were con-

THORNned to one section of the THORNeld (two beds wide and100 feet long) At the Hartnell and Spreckels sites thetrials were carried out in plots of the same dimensionEntrust treatments were initiated 1 wk after trans-planting or thinning of romaine We began samplingthe next week Plots were sampled once a week untilharvest

On the three organic farms we evaluated naturalinfestations of N ribisnigri and other aphids attackingromaine For the Hartnell and Spreckels trials we trans-planted lettuce seedlings infested with N ribisnigriTrays of lettuce transplants were conTHORNned for 3 d withina laboratory colony of N ribisnigri to establish an infes-tation before transplanting at these two THORNeld trials

Three whole romaine plants were sampled fromeach plot each week for a total of 12 plants per treat-ment per week Plants were placed in plastic bags andtaken to the University of California Cooperative Ex-tension laboratory in Salinas for processing An 189-liter (5-gal) plastic water bottle was modiTHORNed to func-tion as a collection sieve for insects on the romaineThe lower half of the bottle was cut away and thecentral portion of the screw-on cap was replaced withorgandy cloth by using a glue gun The half-bottle wasplaced in a large sink with the cap end down Plantswere washed over the cut-away end of the bottle sothat the water carried all insects into the screw-on capwhere material collected on the organdy Plants werewashed using a hand-held showerhead-type faucetThe cap was removed after the plant had been thor-oughly washed and its contents were examined undera microscope The number and species of aphid wererecorded for each plant as was the number of syrphideggs and larvae the number of parasitized aphids andthe number and type of other predators

Syrphid larvae were collected from samples andplaced individually in a petri dish (100 by 25 cmThermo Fisher ScientiTHORNc Waltham MA) for rearingThe larvae were kept in a growth chamber (modelMB60-B Percival ScientiTHORNc Perry IA) with a photo-period of 168 (LD) h at 20ETH25 C and 60ETH75 RH andthey were provided with N ribisnigri pea aphidandor foxglove aphid on pieces of lettuce leaf or fababeanVicia fabaL leaf every 3 d until pupation Uponreaching the adult stage syrphids were allowed 1 d toexpel abdominalszliguid then theywere frozen and laterpinned and placed in a reference collection The spe-cies of each individual was recorded

To test for differences in whole plant densities ofaphids syrphid larvae nonsyrphid predators and syr-phid eggs between the two treatments over time andamong trials we analyzed together the THORNnal 5 wk ofdata from each trial by using a repeated measuresanalysis of variance (ANOVA) model Fixed factors inthe model were trial and treatment (between subject)and sampling week (within subject) The random ef-fects of block (trial) and treatment block (trial)were included when associated covariance parame-ters were estimated to be greater than zero SigniTHORNcantinteractions between factors were followed with testsof simple effects of treatment within trial or weekwithin trial

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1527

Aphid parasitism was determined by counts ofaphid mummies We tested for differences in parasit-ism between treatments and over time separately foreach trial with a repeated measures ANOVA modelTreatment (between-subject) and sampling week(within-subject) were THORNxed factors in the model andthe random effect of block was included when theassociated covariance parameter was estimated to begreater than zero SigniTHORNcant interactions were fol-lowed with tests of simple effects of treatment withinweek

To examine the numerical relationships betweenaphid and syrphid larva densities and between aphidand nonsyrphid predator densities across all trials weapplied a multiple linear regression model to the fulldata set This model included trial and week as THORNxedcategorical variables syrphid larvae and nonsyrphidpredators as THORNxed continuous variables and interac-tions between trial and each continuous variable

All analyses were performed using the MIXEDprocedure of SAS (SAS Institute 2002ETH2003) whichestimates parameters using restricted maximum like-lihood A THORNrst-order autoregressive covariance modelwas used to incorporate correlations in the data col-lected over time Denominator degrees of freedomwere adjusted using the KenwardETHRogers calculationWhen necessary to meet the assumption of normalitydata were transformed using a natural log transforma-tion although untransformed least squares means arepresented to facilitate interpretation Statistical sig-niTHORNcance for all tests was deTHORNned as 005

Results

Aphid Densities N ribisnigri was the predominantaphid species found in each trial although foxgloveaphid potato aphid and green peach aphid also con-tributed to infestations Whole plant densities ofaphids were higher in plots treated with Entrust thanin untreated plots in each trial usually beginning onthe third or fourth sampling week of a trial (trial treatment week interaction Table 1) The highestaverage whole plant aphid infestation levels in ro-maine treated with Entrust were as much as 36 timeshigher than peak infestation levels in untreated romainein the same trial (Table 2) At-harvest aphid infestationlevels for romaine treated with Entrust ranged from2802 922 (mean SEM) per plant in the Hollistertrial to 14003 922 in the Spreckels trial Romainetreated with Entrust was unmarketable because of aphidinfestation In contrast average aphid densities per plantat harvest in untreated romaine ranged from 219 922in the Hartnell trial to 948 922 at the Spreckels siteUntreated romaine was marketableSyrphid Densities Syrphid larvae composed be-

tween 85 and 96 of predators collected from un-treated romaine in the Chualar Hartnell Hollisterand Watsonville trials and 63 of the predators col-lected from untreated romaine at Spreckels (Table 3)Syrphid larvae were the only predator group recov-ered from untreated romaine at all THORNve sites Usuallyby the third week and always by the THORNnal (harvest)week of sampling for each trial whole plant densitiesof syrphids were signiTHORNcantly higher in untreated plots

Table 1 Results from ANOVA comparing whole plant densities of aphids syrphid larvae syrphid eggs and nonsyrphid predatorsbetween untreated plots and plots treated with Entrust and across trials and time

Factors and sources of variationdf

F statistic P valueNumerator Denominator

Aphid densityTrial 4 301 3700 00001Treatment 1 301 24844 00001Trial treatment 4 301 444 00062Sampling wk 4 218 3384 00001Treatment sampling wk 4 218 5910 00001Trial treatment sampling wk 32 218 1018 00001

Syrphid densityTrial 4 15 807 00011Treatment 1 15 28336 00001Trial treatment 4 15 646 00031Sampling wk 4 222 8725 00001Treatment sampling wk 4 222 3751 00001Trial treatment sampling wk 32 222 1098 00001

Syrphid egg densityTrial 4 15 3153 00001Treatment 1 534 1718 00001Trial treatment 4 534 168 01539Sampling wk 4 534 4879 00001Treatment sampling wk 4 534 604 00001Trial treatment sampling wk 32 534 1203 00001

Nonsyrphid predator densityTrial 4 136 323 00144Treatment 1 136 869 00038Trial treatment 4 136 851 00001Sampling wk 4 281 3284 00001Treatment sampling wk 4 281 621 00001Trial treatment sampling wk 32 283 223 00003

1528 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

compared with Entrust plots (trial treatment week interaction Table 1) The highest average wholeplant syrphid densities in untreated romaine rangedfrom 275 058 (at Spreckels during the last week)to 908 058 (at Hartnell in the fourth week) (Table4) The highest average whole plant syrphid densitiesin romaine treated with Entrust ranged from 017 058 (at Spreckels in the fourth week) to 284 058(at Hartnell also in the fourth week)Syrphid Egg Densities Average whole plant densi-

ties of syrphid eggs differed statistically between treat-ments in three of the THORNve trials but only during certainweeks (trial treatment week interaction Table1) Average whole plant densities of syrphid eggs weresigniTHORNcantly higher in romaine treated with Entrustthan in untreated romaine during weeks 4 and 5 of theWatsonville trial (week 4 Entrust 870 074 un-

treated 282 074 week 5 Entrust 592 074untreated 053 074) and during the last samplingweek at the Hartnell trial (Entrust 184 074untreated 042 074) Syrphid egg densities werehigher in untreated romaine during the third week atthe Spreckels trial (Entrust 008 074 untreated 125 074) Whole plant densities of syrphid eggsaveraged 246 026 at the Chualar trial and 297 026 at the Hollister trialSyrphid Species Eight syrphid species and one syr-

phid parasitoid species (Diplazon sp Ichneu-monidae) were reared from larvae collected duringthe course of these trials Each of these species hadbeen recovered during a survey of syrphids in organicromaine in 2005 (Smith and Chaney 2007) with theexception of Sphaerophoria contigua Maquart Spha-erophoria sulfuripes (Thomson) Allograpta obliqua

Table 2 Simple effect test results and untransformed least squares means SEM for whole plant aphid densities between treatmentswithin each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 1679 922 5001 922 11613 922 4355 922 5285 922Untreated 1593 922 1653 922 2235 922 242 922 247 922F P value 056 0456 1683 00001 3382 00001 5961 00001 6012 00001

HollisterEntrust 2367 922 6301 922 6529 922 6378 922 2802 922Untreated 1912 922 6977 922 6097 922 072 922 605 924F P value 003 0859 002 0877 000 0992 9515 00001 1227 0001

HartnellEntrust 441 922 1038 922 2162 922 2864 922 7985 922Untreated 148 922 543 922 1456 922 136 922 219 922F P value 351 0061 301 0083 883 0003 4955 00001 10025 00001

WatsonvilleEntrust 4271 922 5135 922 6425 922 16119 922 6424 922Untreated 4042 922 3334 922 3298 922 871 922 570 922F P value 003 0871 315 0077 399 0046 6715 00001 5814 00001

SpreckelsEntrust 580 922 954 922 3951 922 9684 922 14003 922Untreated 243 922 1663 922 5074 922 2389 922 948 922F P value 243 0120 003 0853 003 0867 1997 00001 6559 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

Table 3 Abundance of all predators and percentages of predator groups in each treatment by trial

TrialTotalno

total no

Syrphidlarvae

Dwarf spiderMinute pirate

bugLadybird beetle Lacewing Rove beetle Bigeyed bug

ChualarUntreated 258 85 7 3 Ntildea Ntilde 4 NtildeEntrust 16 81 13 6 Ntilde Ntilde Ntilde Ntilde

HollisterUntreated 189 88 Ntilde Ntilde 10 2 Ntilde NtildeEntrust 77 47 1 Ntilde 34 18 Ntilde Ntilde

HartnellUntreated 233 96 3 Ntilde 1 1 Ntilde NtildeEntrust 65 80 8 Ntilde 8 4 Ntilde Ntilde

WatsonvilleUntreated 229 91 2 1 Ntilde 1 6 NtildeEntrust 50 40 10 8 4 6 32 Ntilde

SpreckelsUntreated 109 63 2 30 4 1 Ntilde NtildeEntrust 17 18 Ntilde 12 53 12 Ntilde 6

aNtilde not collected

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1529

(Say) Toxomerus marginatus (Say) and Allograptaexotica Wiedemann were the most common speciesreared Syrphus opinatorOsten SackenPlatycheirus steg-nus (Say) Eupeodes americanusWiedemann and Spha-erophoria continua were recovered in low numbersNonsyrphid Predators Whole-plant densities of

nonsyrphid predators generally were low in all trialsAverage whole plant densities of nonsyrphid preda-tors ranged from zero in multiple weeks to 174 021in romaine treated with Entrust during the THORNnal weekof sampling at the Hollister site Entrust affected non-syrphid predators differently depending on trial andsampling week (trial treatment week interaction

Table 1) At the Chualar and Spreckels trials nonsyr-phid predator densities were signiTHORNcantly higher inthe untreated lettuce compared with romaine treatedwith Entrust for the last 3 and 2 weeks of samplingrespectively (Table 5) In contrast at the Hollister andWatsonville trials nonsyrphid predators were signif-icantly higher in romaine treated with Entrust than inuntreated romaine during the week before harvestNonsyrphid predators did not differ statistically be-tween treatments at the Hartnell trial (Table 5)

Nonsyrphid predators collected from organic ro-maine included dwarf spiders (Linyphiidae) minutepirate bugs (Orius sp) ladybird beetles (Coccineli-

Table 4 Simple effect test results and untransformed least squares means SEM for whole plant syrphid larva densities betweentreatments within each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 000 058 050 058 025 058 033 058 000 058Untreated 025 058 329 058 896 058 350 058 232 058F P value 044 0509 2942 00001 12734 00001 3569 00001 2408 00001

HollisterEntrust 000 058 000 058 009 058 073 058 219 058Untreated 000 058 000 058 076 058 649 058 666 058F P value 000 1000 000 0996 240 00001 6562 00001 3758 00001

HartnellEntrust 009 058 000 058 125 060 284 058 025 058Untreated 000 058 066 058 286 058 908 058 598 058F P value 005 0825 285 0094 868 0004 4431 00001 8076 00001

WatsonvilleEntrust 017 058 042 058 025 058 041 058 042 058Untreated 076 058 281 058 530 058 546 058 334 058F P value 211 0149 1922 00001 4336 00001 6067 00001 3255 00001

SpreckelsEntrust 000 058 000 058 000 058 017 058 008 058Untreated 008 058 026 058 100 058 168 058 275 058F P value 005 0821 044 0510 433 0040 113 0001 2904 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

Table 5 Simple effect test results and untransformed least squares means SEM for whole plant nonsyrphid predator densitiesbetween treatments within each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 007 021 000 021 000 021 001 021 016 021Untreated 010 021 016 021 063 021 108 021 116 021F P value 002 0893 076 0384 1001 0002 2123 00001 745 0007

HollisterEntrust 000 021 000 021 034 021 134 021 174 021Untreated 000 021 001 021 044 021 089 021 051 021F P value 000 0994 001 0910 000 0991 042 0516 1744 00001

HartnellEntrust 000 021 009 021 000 021 007 021 093 021Untreated 000 021 000 021 000 021 041 021 045 021F P value 000 0952 027 0610 000 0945 166 0199 316 0077

WatsonvilleEntrust 024 021 000 021 000 021 025 021 125 021Untreated 022 021 002 021 016 021 065 021 036 021F P value 057 0450 001 0920 082 0365 147 0226 622 0013

SpreckelsEntrust 000 021 016 021 001 021 000 021 000 021Untreated 000 021 000 021 040 021 103 021 134 021F P value 000 0974 091 0340 252 0114 1592 00001 2281 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

1530 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

dae) brown lacewings (Hemerobiidae) rove beetles(Staphylinidae) and big-eyed bugs (Geocoris sp)The complex of nonsyrphid predators was different ateach site (Table 3)ParasitismAphid mummies were found only in the

Hartnell and Watsonville trials Average whole plantdensities of aphid mummies were higher in romainetreated with Entrust (143 022) than in untreatedromaine (008 022) at the Hartnell trial (F 1871df 1 6 P 0005) At the Watsonville trial averagewhole plant densities of aphid mummies were signif-icantly higher in romaine treated with Entrust only inweek 5 (Entrust 325 049 untreated 020 049 F 2113 df 1 6 P 0004)Aphid Density Relationships with Syrphid Larvaeand Nonsyrphid Predators Regression analysis re-vealed a signiTHORNcant negative relationship betweenaphid and syrphid densities in all trials but a signiTHORNcantnegativerelationshipbetweennonsyrphidpredatorsandaphiddensities in theChualar trialonly(Table6)Of thefourothertrialswefoundnorelationshipbetweenaphidand nonsyrphid predator densities

Discussion

On CaliforniaOtildes Central Coast several differenttypes of predators are found in organically grownromaine that is infested with N ribisnigri and otheraphids The data presented here and previously(Smith and Chaney 2007) indicate that syrphid larvaeare the most abundant predators often making up85 of the predators collected from infested ro-maine and that they can average up to nine larvae perromaine head during the crucial weeks immediatelybefore the harvest of the crop In addition syrphidlarvae are the only predator group consistently foundin all infested romaine THORNelds Other predators are lessabundant and they are recovered in some THORNelds butnot others Regression analysis presented here sug-gests that increases in syrphid larvae help explainreductions in aphid numbers and that there is no

relationship between nonsyrphid predators and aphiddensities at most sites

We present data that demonstrates that aphid den-sities were signiTHORNcantly higher in plots where syrphidswere suppressed Romaine in which syrphids weresuppressed was unmarketable because of aphid infes-tation whereas untreated romaine was marketableOur results suggest that syrphids play a major role inthe conservation biological control ofN ribisnigri andother aphids in organically grown romaine on theCentral Coast of California

We found that although Entrust does not affectoviposition by syrphid females it does suppress non-syrphid predators in some instances Because Entrustapparently reduced the densities of nonsyrphid pred-ators as well as syrphid larvae at the Chualar andSpreckels trials it was an imperfect tool for demon-strating the effect of selectively removing syrphid lar-vae from organically grown romaine Given thatminute pirate bugs composed 30 of the predators atthe Spreckels trial in untreated romaine and only 12where Entrust was applied it is possible that suppres-sion of minute pirate bugs contributed to the releaseof aphids from predation in the Entrust treatmentDwarf spiders and minute pirate bugs in the Chualartrial also may have contributed to aphid suppressionDwarf spiders are important predators of aphids inwheat Triticum aestivum L (Harwood et al 2004Nienstedt and Poehling 2004) and minute pirate bugsare important predators of soybean aphid (Costama-gna and Landis 2007)

The Chualar and Spreckels trials illustrate instancesinwhichpredatorsother thansyrphidsmaycontributesigniTHORNcantly to the suppression of aphids in organicromaine However unlike the Chualar site whereaverage syrphid densities were among the highest ofall trials 2 wk before harvest syrphid densities werevery low at Spreckels The Spreckels trial also wasunusual in that the aphid infestation in the untreatedplots at harvest approached unmarketable levels Con-sidering the paucity of syrphid larvae the abundance

Table 6 Multiple linear regression analysis and parameter estimates SE for relationships between aphid (y) and syrphid larva (x)densities and between aphid (y) and nonsyrphid predator (z) densities

Factors and sources ofvariation

dfF statistic P value

Numerator Denominator

Trial 4 864 561 0001Syrphid Larvae 1 542 1792 00001Nonsyrphid predators 1 493 000 0982Syrphid larvae trial 4 567 142 0226Nonsyrphid predators trial 4 512 260 0035Sampling wk 4 467 1015 00001

SigniTHORNcance of slope parameter

Parameter estimates DF t value P value

Syrphid larvay 288(009) 0091(002)x 542 423 00001Nonsyrphid predator trialChualar y 298(021) 0302(011)z 505 278 0006Hollister y 330(021) 0080(009)z 486 090 0371Hartnell y 218(021) 0018(017)z 487 011 0914Watsonville y 349(022) 0018(012)z 510 015 0882Spreckels y 288(021) 0179(013)z 546 134 0181

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1531

of minute pirate bugs and the poor biological controlof aphids achieved at this site the Spreckels trial mayoffer an example of interguild predation interferingwith suppression of aphids by syrphids in organicromaine Interguild predation interferes with conser-vation biological control in some instances (Rosen-heim 1998) and further study is required to determinewhether the presence of certain nonsyrphid predatorsreduces the efTHORNcacy of syrphids in suppressing aphidsin organic romaine

It is worth noting that nonsyrphid predators werestatistically higher in romaine treated with Entrust inthe Hollister and Watsonville trials but that Entrust-treated romaine was still unmarketable in these trialsbecause of aphid infestation Numbers of aphid mum-mies also were higher in romaine treated with Entrustin two trials presumably because the densities ofaphids were much higher in romaine treated withEntrust

Given the abundance of syrphid larvae and thatsyrphid larvae are the only predator consistently re-covered from organically grown romaine we arguethat syrphid larvae are primarily responsible for sup-pressing N ribisnigri and other aphids to below eco-nomic levels

Acknowledgments

We thank Robert Bugg (University of California Sustain-able Agriculture Research and Development Program) andDiana Henderson Reisa Bigelow Franklin Dlott Cathy Carl-son Lourdes Rodriguez and Jason Hoeksema for assistancewith this research Sanford Eigenbrode and three anonymousreviewers helped improve the original manuscript We alsothank the farm managers who collaborated with on-farmtrials This research was supported in part by the UnitedStates Environmental Protection Agency Region 9

References Cited

Arthurs S P L A Lacey and E R Miliczky 2007 Eval-uation of codling moth granulosis virus and spinosad forcodling moth control and impact on non-target species inpear orchards Biol Control 41 99ETH109

California Farm Bureau 2006 California Farm Bureau(wwwcfbfcominfoagfacftscfm)

Chaney W E 1999 Lettuce aphid update MontereyCounty Crop Notes University of California CooperativeExtension Salinas CA

Colfer R 2004 Using habitat management to improvebiological control on commercial organic farms in Cal-

ifornia pp 55ETH62 In M S Hoddle [ed] Fourth Cal-ifornia Conference on Biological Control 13ETH15 July2004 Berkeley CA University of California PressBerkeley CA

Contreras J V Quinto V Abellan J Fernandez E Grava-los C Moros and P Bielza 2006 Effects of naturalinsecticides on Frankliniella occidentalis and Orius sppBull OILBSROP 29 331ETH336

Costamagna A C andDA Landis 2007 Quantifying pre-dation on soybean aphid through direct THORNeld observa-tions Biol Control 42 16ETH24

Eizen G W P J Elzen and E G King 1998 Laboratorytoxicity of insecticide residues to Orius insidiosus Geo-coris punctipes Hippodamia convergens and Chrysoperlacarneapp 1235ETH1238 InP Dugger and D Richter [eds]1998 Proceedings of the Beltwide Cotton Conference5ETH9 January 1998 San Diego CA

Harwood J D K D Sunderland and WOC Symondson2004 Prey selection by linyphiid spiders moleculartracking of the effects of alternative prey on rates of aphidconsumption in the THORNeld Mol Ecol 13 3549ETH3560

Klonsky K and K Richter 2007 Statistical review of Cal-ifornia Organic Agriculture 2000ETH2005 Agricultural Is-sues Center University of California Davis CA (aicuc-davisedu)

Liu Y-B 2004 Distribution and population developmentof Nasonovia ribisnigri in iceberg lettuce J Econ Ento-mol 97 883ETH890

MacKenzie J R and R S Vernon 1988 Sampling for thedistribution of lettuce aphid Nasonovia ribisnigri (Ho-moptera Aphididae) in THORNelds and within heads J En-tomol Soc Br Columbia 85 10ETH14

Miles M 2006 The effects of spinosad on beneTHORNcial insectsand mites used in integrated pest management systems ingreenhouses Bull OILSROP 29 53ETH59 [NASS] Na-tional Agricultural Statistics Service 2006 National Ag-ricultural Statistics Service (wwwnassusdagov)

Nienstedt K M and H-M Poehling 2004 Invertebratepredation of 15N-marked prey in semi-THORNeld wheat enclo-sures Entomol Exp Appl 112 191ETH200

Rosenheim J A 1998 High order predators and the regu-lation of insect herbivore populations Annu Rev Ento-mol 43 421ETH47

SAS Insitute 2002ndash2003 SAS version 91 SAS InstituteCary NC

Smith H A and W E Chaney 2007 A survey of syrphidpredators ofNasonovia ribisnigri in organic lettuce on theCentral Coast of California J Econ Entomol 100 39ETH48

Studebaker G E and T J Kring 2003 Effects of insecti-cides on Orius insidiosus measured by THORNeld greenhouseand Petri dish bioassays Fla Entomol 86 178ETH185

Received 21 August 2007 accepted 27 April 2008

1532 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

Aphid parasitism was determined by counts ofaphid mummies We tested for differences in parasit-ism between treatments and over time separately foreach trial with a repeated measures ANOVA modelTreatment (between-subject) and sampling week(within-subject) were THORNxed factors in the model andthe random effect of block was included when theassociated covariance parameter was estimated to begreater than zero SigniTHORNcant interactions were fol-lowed with tests of simple effects of treatment withinweek

To examine the numerical relationships betweenaphid and syrphid larva densities and between aphidand nonsyrphid predator densities across all trials weapplied a multiple linear regression model to the fulldata set This model included trial and week as THORNxedcategorical variables syrphid larvae and nonsyrphidpredators as THORNxed continuous variables and interac-tions between trial and each continuous variable

All analyses were performed using the MIXEDprocedure of SAS (SAS Institute 2002ETH2003) whichestimates parameters using restricted maximum like-lihood A THORNrst-order autoregressive covariance modelwas used to incorporate correlations in the data col-lected over time Denominator degrees of freedomwere adjusted using the KenwardETHRogers calculationWhen necessary to meet the assumption of normalitydata were transformed using a natural log transforma-tion although untransformed least squares means arepresented to facilitate interpretation Statistical sig-niTHORNcance for all tests was deTHORNned as 005

Results

Aphid Densities N ribisnigri was the predominantaphid species found in each trial although foxgloveaphid potato aphid and green peach aphid also con-tributed to infestations Whole plant densities ofaphids were higher in plots treated with Entrust thanin untreated plots in each trial usually beginning onthe third or fourth sampling week of a trial (trial treatment week interaction Table 1) The highestaverage whole plant aphid infestation levels in ro-maine treated with Entrust were as much as 36 timeshigher than peak infestation levels in untreated romainein the same trial (Table 2) At-harvest aphid infestationlevels for romaine treated with Entrust ranged from2802 922 (mean SEM) per plant in the Hollistertrial to 14003 922 in the Spreckels trial Romainetreated with Entrust was unmarketable because of aphidinfestation In contrast average aphid densities per plantat harvest in untreated romaine ranged from 219 922in the Hartnell trial to 948 922 at the Spreckels siteUntreated romaine was marketableSyrphid Densities Syrphid larvae composed be-

tween 85 and 96 of predators collected from un-treated romaine in the Chualar Hartnell Hollisterand Watsonville trials and 63 of the predators col-lected from untreated romaine at Spreckels (Table 3)Syrphid larvae were the only predator group recov-ered from untreated romaine at all THORNve sites Usuallyby the third week and always by the THORNnal (harvest)week of sampling for each trial whole plant densitiesof syrphids were signiTHORNcantly higher in untreated plots

Table 1 Results from ANOVA comparing whole plant densities of aphids syrphid larvae syrphid eggs and nonsyrphid predatorsbetween untreated plots and plots treated with Entrust and across trials and time

Factors and sources of variationdf

F statistic P valueNumerator Denominator

Aphid densityTrial 4 301 3700 00001Treatment 1 301 24844 00001Trial treatment 4 301 444 00062Sampling wk 4 218 3384 00001Treatment sampling wk 4 218 5910 00001Trial treatment sampling wk 32 218 1018 00001

Syrphid densityTrial 4 15 807 00011Treatment 1 15 28336 00001Trial treatment 4 15 646 00031Sampling wk 4 222 8725 00001Treatment sampling wk 4 222 3751 00001Trial treatment sampling wk 32 222 1098 00001

Syrphid egg densityTrial 4 15 3153 00001Treatment 1 534 1718 00001Trial treatment 4 534 168 01539Sampling wk 4 534 4879 00001Treatment sampling wk 4 534 604 00001Trial treatment sampling wk 32 534 1203 00001

Nonsyrphid predator densityTrial 4 136 323 00144Treatment 1 136 869 00038Trial treatment 4 136 851 00001Sampling wk 4 281 3284 00001Treatment sampling wk 4 281 621 00001Trial treatment sampling wk 32 283 223 00003

1528 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

compared with Entrust plots (trial treatment week interaction Table 1) The highest average wholeplant syrphid densities in untreated romaine rangedfrom 275 058 (at Spreckels during the last week)to 908 058 (at Hartnell in the fourth week) (Table4) The highest average whole plant syrphid densitiesin romaine treated with Entrust ranged from 017 058 (at Spreckels in the fourth week) to 284 058(at Hartnell also in the fourth week)Syrphid Egg Densities Average whole plant densi-

ties of syrphid eggs differed statistically between treat-ments in three of the THORNve trials but only during certainweeks (trial treatment week interaction Table1) Average whole plant densities of syrphid eggs weresigniTHORNcantly higher in romaine treated with Entrustthan in untreated romaine during weeks 4 and 5 of theWatsonville trial (week 4 Entrust 870 074 un-

treated 282 074 week 5 Entrust 592 074untreated 053 074) and during the last samplingweek at the Hartnell trial (Entrust 184 074untreated 042 074) Syrphid egg densities werehigher in untreated romaine during the third week atthe Spreckels trial (Entrust 008 074 untreated 125 074) Whole plant densities of syrphid eggsaveraged 246 026 at the Chualar trial and 297 026 at the Hollister trialSyrphid Species Eight syrphid species and one syr-

phid parasitoid species (Diplazon sp Ichneu-monidae) were reared from larvae collected duringthe course of these trials Each of these species hadbeen recovered during a survey of syrphids in organicromaine in 2005 (Smith and Chaney 2007) with theexception of Sphaerophoria contigua Maquart Spha-erophoria sulfuripes (Thomson) Allograpta obliqua

Table 2 Simple effect test results and untransformed least squares means SEM for whole plant aphid densities between treatmentswithin each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 1679 922 5001 922 11613 922 4355 922 5285 922Untreated 1593 922 1653 922 2235 922 242 922 247 922F P value 056 0456 1683 00001 3382 00001 5961 00001 6012 00001

HollisterEntrust 2367 922 6301 922 6529 922 6378 922 2802 922Untreated 1912 922 6977 922 6097 922 072 922 605 924F P value 003 0859 002 0877 000 0992 9515 00001 1227 0001

HartnellEntrust 441 922 1038 922 2162 922 2864 922 7985 922Untreated 148 922 543 922 1456 922 136 922 219 922F P value 351 0061 301 0083 883 0003 4955 00001 10025 00001

WatsonvilleEntrust 4271 922 5135 922 6425 922 16119 922 6424 922Untreated 4042 922 3334 922 3298 922 871 922 570 922F P value 003 0871 315 0077 399 0046 6715 00001 5814 00001

SpreckelsEntrust 580 922 954 922 3951 922 9684 922 14003 922Untreated 243 922 1663 922 5074 922 2389 922 948 922F P value 243 0120 003 0853 003 0867 1997 00001 6559 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

Table 3 Abundance of all predators and percentages of predator groups in each treatment by trial

TrialTotalno

total no

Syrphidlarvae

Dwarf spiderMinute pirate

bugLadybird beetle Lacewing Rove beetle Bigeyed bug

ChualarUntreated 258 85 7 3 Ntildea Ntilde 4 NtildeEntrust 16 81 13 6 Ntilde Ntilde Ntilde Ntilde

HollisterUntreated 189 88 Ntilde Ntilde 10 2 Ntilde NtildeEntrust 77 47 1 Ntilde 34 18 Ntilde Ntilde

HartnellUntreated 233 96 3 Ntilde 1 1 Ntilde NtildeEntrust 65 80 8 Ntilde 8 4 Ntilde Ntilde

WatsonvilleUntreated 229 91 2 1 Ntilde 1 6 NtildeEntrust 50 40 10 8 4 6 32 Ntilde

SpreckelsUntreated 109 63 2 30 4 1 Ntilde NtildeEntrust 17 18 Ntilde 12 53 12 Ntilde 6

aNtilde not collected

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1529

(Say) Toxomerus marginatus (Say) and Allograptaexotica Wiedemann were the most common speciesreared Syrphus opinatorOsten SackenPlatycheirus steg-nus (Say) Eupeodes americanusWiedemann and Spha-erophoria continua were recovered in low numbersNonsyrphid Predators Whole-plant densities of

nonsyrphid predators generally were low in all trialsAverage whole plant densities of nonsyrphid preda-tors ranged from zero in multiple weeks to 174 021in romaine treated with Entrust during the THORNnal weekof sampling at the Hollister site Entrust affected non-syrphid predators differently depending on trial andsampling week (trial treatment week interaction

Table 1) At the Chualar and Spreckels trials nonsyr-phid predator densities were signiTHORNcantly higher inthe untreated lettuce compared with romaine treatedwith Entrust for the last 3 and 2 weeks of samplingrespectively (Table 5) In contrast at the Hollister andWatsonville trials nonsyrphid predators were signif-icantly higher in romaine treated with Entrust than inuntreated romaine during the week before harvestNonsyrphid predators did not differ statistically be-tween treatments at the Hartnell trial (Table 5)

Nonsyrphid predators collected from organic ro-maine included dwarf spiders (Linyphiidae) minutepirate bugs (Orius sp) ladybird beetles (Coccineli-

Table 4 Simple effect test results and untransformed least squares means SEM for whole plant syrphid larva densities betweentreatments within each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 000 058 050 058 025 058 033 058 000 058Untreated 025 058 329 058 896 058 350 058 232 058F P value 044 0509 2942 00001 12734 00001 3569 00001 2408 00001

HollisterEntrust 000 058 000 058 009 058 073 058 219 058Untreated 000 058 000 058 076 058 649 058 666 058F P value 000 1000 000 0996 240 00001 6562 00001 3758 00001

HartnellEntrust 009 058 000 058 125 060 284 058 025 058Untreated 000 058 066 058 286 058 908 058 598 058F P value 005 0825 285 0094 868 0004 4431 00001 8076 00001

WatsonvilleEntrust 017 058 042 058 025 058 041 058 042 058Untreated 076 058 281 058 530 058 546 058 334 058F P value 211 0149 1922 00001 4336 00001 6067 00001 3255 00001

SpreckelsEntrust 000 058 000 058 000 058 017 058 008 058Untreated 008 058 026 058 100 058 168 058 275 058F P value 005 0821 044 0510 433 0040 113 0001 2904 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

Table 5 Simple effect test results and untransformed least squares means SEM for whole plant nonsyrphid predator densitiesbetween treatments within each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 007 021 000 021 000 021 001 021 016 021Untreated 010 021 016 021 063 021 108 021 116 021F P value 002 0893 076 0384 1001 0002 2123 00001 745 0007

HollisterEntrust 000 021 000 021 034 021 134 021 174 021Untreated 000 021 001 021 044 021 089 021 051 021F P value 000 0994 001 0910 000 0991 042 0516 1744 00001

HartnellEntrust 000 021 009 021 000 021 007 021 093 021Untreated 000 021 000 021 000 021 041 021 045 021F P value 000 0952 027 0610 000 0945 166 0199 316 0077

WatsonvilleEntrust 024 021 000 021 000 021 025 021 125 021Untreated 022 021 002 021 016 021 065 021 036 021F P value 057 0450 001 0920 082 0365 147 0226 622 0013

SpreckelsEntrust 000 021 016 021 001 021 000 021 000 021Untreated 000 021 000 021 040 021 103 021 134 021F P value 000 0974 091 0340 252 0114 1592 00001 2281 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

1530 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

dae) brown lacewings (Hemerobiidae) rove beetles(Staphylinidae) and big-eyed bugs (Geocoris sp)The complex of nonsyrphid predators was different ateach site (Table 3)ParasitismAphid mummies were found only in the

Hartnell and Watsonville trials Average whole plantdensities of aphid mummies were higher in romainetreated with Entrust (143 022) than in untreatedromaine (008 022) at the Hartnell trial (F 1871df 1 6 P 0005) At the Watsonville trial averagewhole plant densities of aphid mummies were signif-icantly higher in romaine treated with Entrust only inweek 5 (Entrust 325 049 untreated 020 049 F 2113 df 1 6 P 0004)Aphid Density Relationships with Syrphid Larvaeand Nonsyrphid Predators Regression analysis re-vealed a signiTHORNcant negative relationship betweenaphid and syrphid densities in all trials but a signiTHORNcantnegativerelationshipbetweennonsyrphidpredatorsandaphiddensities in theChualar trialonly(Table6)Of thefourothertrialswefoundnorelationshipbetweenaphidand nonsyrphid predator densities

Discussion

On CaliforniaOtildes Central Coast several differenttypes of predators are found in organically grownromaine that is infested with N ribisnigri and otheraphids The data presented here and previously(Smith and Chaney 2007) indicate that syrphid larvaeare the most abundant predators often making up85 of the predators collected from infested ro-maine and that they can average up to nine larvae perromaine head during the crucial weeks immediatelybefore the harvest of the crop In addition syrphidlarvae are the only predator group consistently foundin all infested romaine THORNelds Other predators are lessabundant and they are recovered in some THORNelds butnot others Regression analysis presented here sug-gests that increases in syrphid larvae help explainreductions in aphid numbers and that there is no

relationship between nonsyrphid predators and aphiddensities at most sites

We present data that demonstrates that aphid den-sities were signiTHORNcantly higher in plots where syrphidswere suppressed Romaine in which syrphids weresuppressed was unmarketable because of aphid infes-tation whereas untreated romaine was marketableOur results suggest that syrphids play a major role inthe conservation biological control ofN ribisnigri andother aphids in organically grown romaine on theCentral Coast of California

We found that although Entrust does not affectoviposition by syrphid females it does suppress non-syrphid predators in some instances Because Entrustapparently reduced the densities of nonsyrphid pred-ators as well as syrphid larvae at the Chualar andSpreckels trials it was an imperfect tool for demon-strating the effect of selectively removing syrphid lar-vae from organically grown romaine Given thatminute pirate bugs composed 30 of the predators atthe Spreckels trial in untreated romaine and only 12where Entrust was applied it is possible that suppres-sion of minute pirate bugs contributed to the releaseof aphids from predation in the Entrust treatmentDwarf spiders and minute pirate bugs in the Chualartrial also may have contributed to aphid suppressionDwarf spiders are important predators of aphids inwheat Triticum aestivum L (Harwood et al 2004Nienstedt and Poehling 2004) and minute pirate bugsare important predators of soybean aphid (Costama-gna and Landis 2007)

The Chualar and Spreckels trials illustrate instancesinwhichpredatorsother thansyrphidsmaycontributesigniTHORNcantly to the suppression of aphids in organicromaine However unlike the Chualar site whereaverage syrphid densities were among the highest ofall trials 2 wk before harvest syrphid densities werevery low at Spreckels The Spreckels trial also wasunusual in that the aphid infestation in the untreatedplots at harvest approached unmarketable levels Con-sidering the paucity of syrphid larvae the abundance

Table 6 Multiple linear regression analysis and parameter estimates SE for relationships between aphid (y) and syrphid larva (x)densities and between aphid (y) and nonsyrphid predator (z) densities

Factors and sources ofvariation

dfF statistic P value

Numerator Denominator

Trial 4 864 561 0001Syrphid Larvae 1 542 1792 00001Nonsyrphid predators 1 493 000 0982Syrphid larvae trial 4 567 142 0226Nonsyrphid predators trial 4 512 260 0035Sampling wk 4 467 1015 00001

SigniTHORNcance of slope parameter

Parameter estimates DF t value P value

Syrphid larvay 288(009) 0091(002)x 542 423 00001Nonsyrphid predator trialChualar y 298(021) 0302(011)z 505 278 0006Hollister y 330(021) 0080(009)z 486 090 0371Hartnell y 218(021) 0018(017)z 487 011 0914Watsonville y 349(022) 0018(012)z 510 015 0882Spreckels y 288(021) 0179(013)z 546 134 0181

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1531

of minute pirate bugs and the poor biological controlof aphids achieved at this site the Spreckels trial mayoffer an example of interguild predation interferingwith suppression of aphids by syrphids in organicromaine Interguild predation interferes with conser-vation biological control in some instances (Rosen-heim 1998) and further study is required to determinewhether the presence of certain nonsyrphid predatorsreduces the efTHORNcacy of syrphids in suppressing aphidsin organic romaine

It is worth noting that nonsyrphid predators werestatistically higher in romaine treated with Entrust inthe Hollister and Watsonville trials but that Entrust-treated romaine was still unmarketable in these trialsbecause of aphid infestation Numbers of aphid mum-mies also were higher in romaine treated with Entrustin two trials presumably because the densities ofaphids were much higher in romaine treated withEntrust

Given the abundance of syrphid larvae and thatsyrphid larvae are the only predator consistently re-covered from organically grown romaine we arguethat syrphid larvae are primarily responsible for sup-pressing N ribisnigri and other aphids to below eco-nomic levels

Acknowledgments

We thank Robert Bugg (University of California Sustain-able Agriculture Research and Development Program) andDiana Henderson Reisa Bigelow Franklin Dlott Cathy Carl-son Lourdes Rodriguez and Jason Hoeksema for assistancewith this research Sanford Eigenbrode and three anonymousreviewers helped improve the original manuscript We alsothank the farm managers who collaborated with on-farmtrials This research was supported in part by the UnitedStates Environmental Protection Agency Region 9

References Cited

Arthurs S P L A Lacey and E R Miliczky 2007 Eval-uation of codling moth granulosis virus and spinosad forcodling moth control and impact on non-target species inpear orchards Biol Control 41 99ETH109

California Farm Bureau 2006 California Farm Bureau(wwwcfbfcominfoagfacftscfm)

Chaney W E 1999 Lettuce aphid update MontereyCounty Crop Notes University of California CooperativeExtension Salinas CA

Colfer R 2004 Using habitat management to improvebiological control on commercial organic farms in Cal-

ifornia pp 55ETH62 In M S Hoddle [ed] Fourth Cal-ifornia Conference on Biological Control 13ETH15 July2004 Berkeley CA University of California PressBerkeley CA

Contreras J V Quinto V Abellan J Fernandez E Grava-los C Moros and P Bielza 2006 Effects of naturalinsecticides on Frankliniella occidentalis and Orius sppBull OILBSROP 29 331ETH336

Costamagna A C andDA Landis 2007 Quantifying pre-dation on soybean aphid through direct THORNeld observa-tions Biol Control 42 16ETH24

Eizen G W P J Elzen and E G King 1998 Laboratorytoxicity of insecticide residues to Orius insidiosus Geo-coris punctipes Hippodamia convergens and Chrysoperlacarneapp 1235ETH1238 InP Dugger and D Richter [eds]1998 Proceedings of the Beltwide Cotton Conference5ETH9 January 1998 San Diego CA

Harwood J D K D Sunderland and WOC Symondson2004 Prey selection by linyphiid spiders moleculartracking of the effects of alternative prey on rates of aphidconsumption in the THORNeld Mol Ecol 13 3549ETH3560

Klonsky K and K Richter 2007 Statistical review of Cal-ifornia Organic Agriculture 2000ETH2005 Agricultural Is-sues Center University of California Davis CA (aicuc-davisedu)

Liu Y-B 2004 Distribution and population developmentof Nasonovia ribisnigri in iceberg lettuce J Econ Ento-mol 97 883ETH890

MacKenzie J R and R S Vernon 1988 Sampling for thedistribution of lettuce aphid Nasonovia ribisnigri (Ho-moptera Aphididae) in THORNelds and within heads J En-tomol Soc Br Columbia 85 10ETH14

Miles M 2006 The effects of spinosad on beneTHORNcial insectsand mites used in integrated pest management systems ingreenhouses Bull OILSROP 29 53ETH59 [NASS] Na-tional Agricultural Statistics Service 2006 National Ag-ricultural Statistics Service (wwwnassusdagov)

Nienstedt K M and H-M Poehling 2004 Invertebratepredation of 15N-marked prey in semi-THORNeld wheat enclo-sures Entomol Exp Appl 112 191ETH200

Rosenheim J A 1998 High order predators and the regu-lation of insect herbivore populations Annu Rev Ento-mol 43 421ETH47

SAS Insitute 2002ndash2003 SAS version 91 SAS InstituteCary NC

Smith H A and W E Chaney 2007 A survey of syrphidpredators ofNasonovia ribisnigri in organic lettuce on theCentral Coast of California J Econ Entomol 100 39ETH48

Studebaker G E and T J Kring 2003 Effects of insecti-cides on Orius insidiosus measured by THORNeld greenhouseand Petri dish bioassays Fla Entomol 86 178ETH185

Received 21 August 2007 accepted 27 April 2008

1532 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

compared with Entrust plots (trial treatment week interaction Table 1) The highest average wholeplant syrphid densities in untreated romaine rangedfrom 275 058 (at Spreckels during the last week)to 908 058 (at Hartnell in the fourth week) (Table4) The highest average whole plant syrphid densitiesin romaine treated with Entrust ranged from 017 058 (at Spreckels in the fourth week) to 284 058(at Hartnell also in the fourth week)Syrphid Egg Densities Average whole plant densi-

ties of syrphid eggs differed statistically between treat-ments in three of the THORNve trials but only during certainweeks (trial treatment week interaction Table1) Average whole plant densities of syrphid eggs weresigniTHORNcantly higher in romaine treated with Entrustthan in untreated romaine during weeks 4 and 5 of theWatsonville trial (week 4 Entrust 870 074 un-

treated 282 074 week 5 Entrust 592 074untreated 053 074) and during the last samplingweek at the Hartnell trial (Entrust 184 074untreated 042 074) Syrphid egg densities werehigher in untreated romaine during the third week atthe Spreckels trial (Entrust 008 074 untreated 125 074) Whole plant densities of syrphid eggsaveraged 246 026 at the Chualar trial and 297 026 at the Hollister trialSyrphid Species Eight syrphid species and one syr-

phid parasitoid species (Diplazon sp Ichneu-monidae) were reared from larvae collected duringthe course of these trials Each of these species hadbeen recovered during a survey of syrphids in organicromaine in 2005 (Smith and Chaney 2007) with theexception of Sphaerophoria contigua Maquart Spha-erophoria sulfuripes (Thomson) Allograpta obliqua

Table 2 Simple effect test results and untransformed least squares means SEM for whole plant aphid densities between treatmentswithin each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 1679 922 5001 922 11613 922 4355 922 5285 922Untreated 1593 922 1653 922 2235 922 242 922 247 922F P value 056 0456 1683 00001 3382 00001 5961 00001 6012 00001

HollisterEntrust 2367 922 6301 922 6529 922 6378 922 2802 922Untreated 1912 922 6977 922 6097 922 072 922 605 924F P value 003 0859 002 0877 000 0992 9515 00001 1227 0001

HartnellEntrust 441 922 1038 922 2162 922 2864 922 7985 922Untreated 148 922 543 922 1456 922 136 922 219 922F P value 351 0061 301 0083 883 0003 4955 00001 10025 00001

WatsonvilleEntrust 4271 922 5135 922 6425 922 16119 922 6424 922Untreated 4042 922 3334 922 3298 922 871 922 570 922F P value 003 0871 315 0077 399 0046 6715 00001 5814 00001

SpreckelsEntrust 580 922 954 922 3951 922 9684 922 14003 922Untreated 243 922 1663 922 5074 922 2389 922 948 922F P value 243 0120 003 0853 003 0867 1997 00001 6559 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

Table 3 Abundance of all predators and percentages of predator groups in each treatment by trial

TrialTotalno

total no

Syrphidlarvae

Dwarf spiderMinute pirate

bugLadybird beetle Lacewing Rove beetle Bigeyed bug

ChualarUntreated 258 85 7 3 Ntildea Ntilde 4 NtildeEntrust 16 81 13 6 Ntilde Ntilde Ntilde Ntilde

HollisterUntreated 189 88 Ntilde Ntilde 10 2 Ntilde NtildeEntrust 77 47 1 Ntilde 34 18 Ntilde Ntilde

HartnellUntreated 233 96 3 Ntilde 1 1 Ntilde NtildeEntrust 65 80 8 Ntilde 8 4 Ntilde Ntilde

WatsonvilleUntreated 229 91 2 1 Ntilde 1 6 NtildeEntrust 50 40 10 8 4 6 32 Ntilde

SpreckelsUntreated 109 63 2 30 4 1 Ntilde NtildeEntrust 17 18 Ntilde 12 53 12 Ntilde 6

aNtilde not collected

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1529

(Say) Toxomerus marginatus (Say) and Allograptaexotica Wiedemann were the most common speciesreared Syrphus opinatorOsten SackenPlatycheirus steg-nus (Say) Eupeodes americanusWiedemann and Spha-erophoria continua were recovered in low numbersNonsyrphid Predators Whole-plant densities of

nonsyrphid predators generally were low in all trialsAverage whole plant densities of nonsyrphid preda-tors ranged from zero in multiple weeks to 174 021in romaine treated with Entrust during the THORNnal weekof sampling at the Hollister site Entrust affected non-syrphid predators differently depending on trial andsampling week (trial treatment week interaction

Table 1) At the Chualar and Spreckels trials nonsyr-phid predator densities were signiTHORNcantly higher inthe untreated lettuce compared with romaine treatedwith Entrust for the last 3 and 2 weeks of samplingrespectively (Table 5) In contrast at the Hollister andWatsonville trials nonsyrphid predators were signif-icantly higher in romaine treated with Entrust than inuntreated romaine during the week before harvestNonsyrphid predators did not differ statistically be-tween treatments at the Hartnell trial (Table 5)

Nonsyrphid predators collected from organic ro-maine included dwarf spiders (Linyphiidae) minutepirate bugs (Orius sp) ladybird beetles (Coccineli-

Table 4 Simple effect test results and untransformed least squares means SEM for whole plant syrphid larva densities betweentreatments within each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 000 058 050 058 025 058 033 058 000 058Untreated 025 058 329 058 896 058 350 058 232 058F P value 044 0509 2942 00001 12734 00001 3569 00001 2408 00001

HollisterEntrust 000 058 000 058 009 058 073 058 219 058Untreated 000 058 000 058 076 058 649 058 666 058F P value 000 1000 000 0996 240 00001 6562 00001 3758 00001

HartnellEntrust 009 058 000 058 125 060 284 058 025 058Untreated 000 058 066 058 286 058 908 058 598 058F P value 005 0825 285 0094 868 0004 4431 00001 8076 00001

WatsonvilleEntrust 017 058 042 058 025 058 041 058 042 058Untreated 076 058 281 058 530 058 546 058 334 058F P value 211 0149 1922 00001 4336 00001 6067 00001 3255 00001

SpreckelsEntrust 000 058 000 058 000 058 017 058 008 058Untreated 008 058 026 058 100 058 168 058 275 058F P value 005 0821 044 0510 433 0040 113 0001 2904 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

Table 5 Simple effect test results and untransformed least squares means SEM for whole plant nonsyrphid predator densitiesbetween treatments within each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 007 021 000 021 000 021 001 021 016 021Untreated 010 021 016 021 063 021 108 021 116 021F P value 002 0893 076 0384 1001 0002 2123 00001 745 0007

HollisterEntrust 000 021 000 021 034 021 134 021 174 021Untreated 000 021 001 021 044 021 089 021 051 021F P value 000 0994 001 0910 000 0991 042 0516 1744 00001

HartnellEntrust 000 021 009 021 000 021 007 021 093 021Untreated 000 021 000 021 000 021 041 021 045 021F P value 000 0952 027 0610 000 0945 166 0199 316 0077

WatsonvilleEntrust 024 021 000 021 000 021 025 021 125 021Untreated 022 021 002 021 016 021 065 021 036 021F P value 057 0450 001 0920 082 0365 147 0226 622 0013

SpreckelsEntrust 000 021 016 021 001 021 000 021 000 021Untreated 000 021 000 021 040 021 103 021 134 021F P value 000 0974 091 0340 252 0114 1592 00001 2281 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

1530 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

dae) brown lacewings (Hemerobiidae) rove beetles(Staphylinidae) and big-eyed bugs (Geocoris sp)The complex of nonsyrphid predators was different ateach site (Table 3)ParasitismAphid mummies were found only in the

Hartnell and Watsonville trials Average whole plantdensities of aphid mummies were higher in romainetreated with Entrust (143 022) than in untreatedromaine (008 022) at the Hartnell trial (F 1871df 1 6 P 0005) At the Watsonville trial averagewhole plant densities of aphid mummies were signif-icantly higher in romaine treated with Entrust only inweek 5 (Entrust 325 049 untreated 020 049 F 2113 df 1 6 P 0004)Aphid Density Relationships with Syrphid Larvaeand Nonsyrphid Predators Regression analysis re-vealed a signiTHORNcant negative relationship betweenaphid and syrphid densities in all trials but a signiTHORNcantnegativerelationshipbetweennonsyrphidpredatorsandaphiddensities in theChualar trialonly(Table6)Of thefourothertrialswefoundnorelationshipbetweenaphidand nonsyrphid predator densities

Discussion

On CaliforniaOtildes Central Coast several differenttypes of predators are found in organically grownromaine that is infested with N ribisnigri and otheraphids The data presented here and previously(Smith and Chaney 2007) indicate that syrphid larvaeare the most abundant predators often making up85 of the predators collected from infested ro-maine and that they can average up to nine larvae perromaine head during the crucial weeks immediatelybefore the harvest of the crop In addition syrphidlarvae are the only predator group consistently foundin all infested romaine THORNelds Other predators are lessabundant and they are recovered in some THORNelds butnot others Regression analysis presented here sug-gests that increases in syrphid larvae help explainreductions in aphid numbers and that there is no

relationship between nonsyrphid predators and aphiddensities at most sites

We present data that demonstrates that aphid den-sities were signiTHORNcantly higher in plots where syrphidswere suppressed Romaine in which syrphids weresuppressed was unmarketable because of aphid infes-tation whereas untreated romaine was marketableOur results suggest that syrphids play a major role inthe conservation biological control ofN ribisnigri andother aphids in organically grown romaine on theCentral Coast of California

We found that although Entrust does not affectoviposition by syrphid females it does suppress non-syrphid predators in some instances Because Entrustapparently reduced the densities of nonsyrphid pred-ators as well as syrphid larvae at the Chualar andSpreckels trials it was an imperfect tool for demon-strating the effect of selectively removing syrphid lar-vae from organically grown romaine Given thatminute pirate bugs composed 30 of the predators atthe Spreckels trial in untreated romaine and only 12where Entrust was applied it is possible that suppres-sion of minute pirate bugs contributed to the releaseof aphids from predation in the Entrust treatmentDwarf spiders and minute pirate bugs in the Chualartrial also may have contributed to aphid suppressionDwarf spiders are important predators of aphids inwheat Triticum aestivum L (Harwood et al 2004Nienstedt and Poehling 2004) and minute pirate bugsare important predators of soybean aphid (Costama-gna and Landis 2007)

The Chualar and Spreckels trials illustrate instancesinwhichpredatorsother thansyrphidsmaycontributesigniTHORNcantly to the suppression of aphids in organicromaine However unlike the Chualar site whereaverage syrphid densities were among the highest ofall trials 2 wk before harvest syrphid densities werevery low at Spreckels The Spreckels trial also wasunusual in that the aphid infestation in the untreatedplots at harvest approached unmarketable levels Con-sidering the paucity of syrphid larvae the abundance

Table 6 Multiple linear regression analysis and parameter estimates SE for relationships between aphid (y) and syrphid larva (x)densities and between aphid (y) and nonsyrphid predator (z) densities

Factors and sources ofvariation

dfF statistic P value

Numerator Denominator

Trial 4 864 561 0001Syrphid Larvae 1 542 1792 00001Nonsyrphid predators 1 493 000 0982Syrphid larvae trial 4 567 142 0226Nonsyrphid predators trial 4 512 260 0035Sampling wk 4 467 1015 00001

SigniTHORNcance of slope parameter

Parameter estimates DF t value P value

Syrphid larvay 288(009) 0091(002)x 542 423 00001Nonsyrphid predator trialChualar y 298(021) 0302(011)z 505 278 0006Hollister y 330(021) 0080(009)z 486 090 0371Hartnell y 218(021) 0018(017)z 487 011 0914Watsonville y 349(022) 0018(012)z 510 015 0882Spreckels y 288(021) 0179(013)z 546 134 0181

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1531

of minute pirate bugs and the poor biological controlof aphids achieved at this site the Spreckels trial mayoffer an example of interguild predation interferingwith suppression of aphids by syrphids in organicromaine Interguild predation interferes with conser-vation biological control in some instances (Rosen-heim 1998) and further study is required to determinewhether the presence of certain nonsyrphid predatorsreduces the efTHORNcacy of syrphids in suppressing aphidsin organic romaine

It is worth noting that nonsyrphid predators werestatistically higher in romaine treated with Entrust inthe Hollister and Watsonville trials but that Entrust-treated romaine was still unmarketable in these trialsbecause of aphid infestation Numbers of aphid mum-mies also were higher in romaine treated with Entrustin two trials presumably because the densities ofaphids were much higher in romaine treated withEntrust

Given the abundance of syrphid larvae and thatsyrphid larvae are the only predator consistently re-covered from organically grown romaine we arguethat syrphid larvae are primarily responsible for sup-pressing N ribisnigri and other aphids to below eco-nomic levels

Acknowledgments

We thank Robert Bugg (University of California Sustain-able Agriculture Research and Development Program) andDiana Henderson Reisa Bigelow Franklin Dlott Cathy Carl-son Lourdes Rodriguez and Jason Hoeksema for assistancewith this research Sanford Eigenbrode and three anonymousreviewers helped improve the original manuscript We alsothank the farm managers who collaborated with on-farmtrials This research was supported in part by the UnitedStates Environmental Protection Agency Region 9

References Cited

Arthurs S P L A Lacey and E R Miliczky 2007 Eval-uation of codling moth granulosis virus and spinosad forcodling moth control and impact on non-target species inpear orchards Biol Control 41 99ETH109

California Farm Bureau 2006 California Farm Bureau(wwwcfbfcominfoagfacftscfm)

Chaney W E 1999 Lettuce aphid update MontereyCounty Crop Notes University of California CooperativeExtension Salinas CA

Colfer R 2004 Using habitat management to improvebiological control on commercial organic farms in Cal-

ifornia pp 55ETH62 In M S Hoddle [ed] Fourth Cal-ifornia Conference on Biological Control 13ETH15 July2004 Berkeley CA University of California PressBerkeley CA

Contreras J V Quinto V Abellan J Fernandez E Grava-los C Moros and P Bielza 2006 Effects of naturalinsecticides on Frankliniella occidentalis and Orius sppBull OILBSROP 29 331ETH336

Costamagna A C andDA Landis 2007 Quantifying pre-dation on soybean aphid through direct THORNeld observa-tions Biol Control 42 16ETH24

Eizen G W P J Elzen and E G King 1998 Laboratorytoxicity of insecticide residues to Orius insidiosus Geo-coris punctipes Hippodamia convergens and Chrysoperlacarneapp 1235ETH1238 InP Dugger and D Richter [eds]1998 Proceedings of the Beltwide Cotton Conference5ETH9 January 1998 San Diego CA

Harwood J D K D Sunderland and WOC Symondson2004 Prey selection by linyphiid spiders moleculartracking of the effects of alternative prey on rates of aphidconsumption in the THORNeld Mol Ecol 13 3549ETH3560

Klonsky K and K Richter 2007 Statistical review of Cal-ifornia Organic Agriculture 2000ETH2005 Agricultural Is-sues Center University of California Davis CA (aicuc-davisedu)

Liu Y-B 2004 Distribution and population developmentof Nasonovia ribisnigri in iceberg lettuce J Econ Ento-mol 97 883ETH890

MacKenzie J R and R S Vernon 1988 Sampling for thedistribution of lettuce aphid Nasonovia ribisnigri (Ho-moptera Aphididae) in THORNelds and within heads J En-tomol Soc Br Columbia 85 10ETH14

Miles M 2006 The effects of spinosad on beneTHORNcial insectsand mites used in integrated pest management systems ingreenhouses Bull OILSROP 29 53ETH59 [NASS] Na-tional Agricultural Statistics Service 2006 National Ag-ricultural Statistics Service (wwwnassusdagov)

Nienstedt K M and H-M Poehling 2004 Invertebratepredation of 15N-marked prey in semi-THORNeld wheat enclo-sures Entomol Exp Appl 112 191ETH200

Rosenheim J A 1998 High order predators and the regu-lation of insect herbivore populations Annu Rev Ento-mol 43 421ETH47

SAS Insitute 2002ndash2003 SAS version 91 SAS InstituteCary NC

Smith H A and W E Chaney 2007 A survey of syrphidpredators ofNasonovia ribisnigri in organic lettuce on theCentral Coast of California J Econ Entomol 100 39ETH48

Studebaker G E and T J Kring 2003 Effects of insecti-cides on Orius insidiosus measured by THORNeld greenhouseand Petri dish bioassays Fla Entomol 86 178ETH185

Received 21 August 2007 accepted 27 April 2008

1532 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

(Say) Toxomerus marginatus (Say) and Allograptaexotica Wiedemann were the most common speciesreared Syrphus opinatorOsten SackenPlatycheirus steg-nus (Say) Eupeodes americanusWiedemann and Spha-erophoria continua were recovered in low numbersNonsyrphid Predators Whole-plant densities of

nonsyrphid predators generally were low in all trialsAverage whole plant densities of nonsyrphid preda-tors ranged from zero in multiple weeks to 174 021in romaine treated with Entrust during the THORNnal weekof sampling at the Hollister site Entrust affected non-syrphid predators differently depending on trial andsampling week (trial treatment week interaction

Table 1) At the Chualar and Spreckels trials nonsyr-phid predator densities were signiTHORNcantly higher inthe untreated lettuce compared with romaine treatedwith Entrust for the last 3 and 2 weeks of samplingrespectively (Table 5) In contrast at the Hollister andWatsonville trials nonsyrphid predators were signif-icantly higher in romaine treated with Entrust than inuntreated romaine during the week before harvestNonsyrphid predators did not differ statistically be-tween treatments at the Hartnell trial (Table 5)

Nonsyrphid predators collected from organic ro-maine included dwarf spiders (Linyphiidae) minutepirate bugs (Orius sp) ladybird beetles (Coccineli-

Table 4 Simple effect test results and untransformed least squares means SEM for whole plant syrphid larva densities betweentreatments within each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 000 058 050 058 025 058 033 058 000 058Untreated 025 058 329 058 896 058 350 058 232 058F P value 044 0509 2942 00001 12734 00001 3569 00001 2408 00001

HollisterEntrust 000 058 000 058 009 058 073 058 219 058Untreated 000 058 000 058 076 058 649 058 666 058F P value 000 1000 000 0996 240 00001 6562 00001 3758 00001

HartnellEntrust 009 058 000 058 125 060 284 058 025 058Untreated 000 058 066 058 286 058 908 058 598 058F P value 005 0825 285 0094 868 0004 4431 00001 8076 00001

WatsonvilleEntrust 017 058 042 058 025 058 041 058 042 058Untreated 076 058 281 058 530 058 546 058 334 058F P value 211 0149 1922 00001 4336 00001 6067 00001 3255 00001

SpreckelsEntrust 000 058 000 058 000 058 017 058 008 058Untreated 008 058 026 058 100 058 168 058 275 058F P value 005 0821 044 0510 433 0040 113 0001 2904 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

Table 5 Simple effect test results and untransformed least squares means SEM for whole plant nonsyrphid predator densitiesbetween treatments within each sampling week at each trial

Sampling wk

1 2 3 4 5

ChualarEntrust 007 021 000 021 000 021 001 021 016 021Untreated 010 021 016 021 063 021 108 021 116 021F P value 002 0893 076 0384 1001 0002 2123 00001 745 0007

HollisterEntrust 000 021 000 021 034 021 134 021 174 021Untreated 000 021 001 021 044 021 089 021 051 021F P value 000 0994 001 0910 000 0991 042 0516 1744 00001

HartnellEntrust 000 021 009 021 000 021 007 021 093 021Untreated 000 021 000 021 000 021 041 021 045 021F P value 000 0952 027 0610 000 0945 166 0199 316 0077

WatsonvilleEntrust 024 021 000 021 000 021 025 021 125 021Untreated 022 021 002 021 016 021 065 021 036 021F P value 057 0450 001 0920 082 0365 147 0226 622 0013

SpreckelsEntrust 000 021 016 021 001 021 000 021 000 021Untreated 000 021 000 021 040 021 103 021 134 021F P value 000 0974 091 0340 252 0114 1592 00001 2281 00001

Statistical differences for pairs of means were determined by 005 with df 1 520

1530 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

dae) brown lacewings (Hemerobiidae) rove beetles(Staphylinidae) and big-eyed bugs (Geocoris sp)The complex of nonsyrphid predators was different ateach site (Table 3)ParasitismAphid mummies were found only in the

Hartnell and Watsonville trials Average whole plantdensities of aphid mummies were higher in romainetreated with Entrust (143 022) than in untreatedromaine (008 022) at the Hartnell trial (F 1871df 1 6 P 0005) At the Watsonville trial averagewhole plant densities of aphid mummies were signif-icantly higher in romaine treated with Entrust only inweek 5 (Entrust 325 049 untreated 020 049 F 2113 df 1 6 P 0004)Aphid Density Relationships with Syrphid Larvaeand Nonsyrphid Predators Regression analysis re-vealed a signiTHORNcant negative relationship betweenaphid and syrphid densities in all trials but a signiTHORNcantnegativerelationshipbetweennonsyrphidpredatorsandaphiddensities in theChualar trialonly(Table6)Of thefourothertrialswefoundnorelationshipbetweenaphidand nonsyrphid predator densities

Discussion

On CaliforniaOtildes Central Coast several differenttypes of predators are found in organically grownromaine that is infested with N ribisnigri and otheraphids The data presented here and previously(Smith and Chaney 2007) indicate that syrphid larvaeare the most abundant predators often making up85 of the predators collected from infested ro-maine and that they can average up to nine larvae perromaine head during the crucial weeks immediatelybefore the harvest of the crop In addition syrphidlarvae are the only predator group consistently foundin all infested romaine THORNelds Other predators are lessabundant and they are recovered in some THORNelds butnot others Regression analysis presented here sug-gests that increases in syrphid larvae help explainreductions in aphid numbers and that there is no

relationship between nonsyrphid predators and aphiddensities at most sites

We present data that demonstrates that aphid den-sities were signiTHORNcantly higher in plots where syrphidswere suppressed Romaine in which syrphids weresuppressed was unmarketable because of aphid infes-tation whereas untreated romaine was marketableOur results suggest that syrphids play a major role inthe conservation biological control ofN ribisnigri andother aphids in organically grown romaine on theCentral Coast of California

We found that although Entrust does not affectoviposition by syrphid females it does suppress non-syrphid predators in some instances Because Entrustapparently reduced the densities of nonsyrphid pred-ators as well as syrphid larvae at the Chualar andSpreckels trials it was an imperfect tool for demon-strating the effect of selectively removing syrphid lar-vae from organically grown romaine Given thatminute pirate bugs composed 30 of the predators atthe Spreckels trial in untreated romaine and only 12where Entrust was applied it is possible that suppres-sion of minute pirate bugs contributed to the releaseof aphids from predation in the Entrust treatmentDwarf spiders and minute pirate bugs in the Chualartrial also may have contributed to aphid suppressionDwarf spiders are important predators of aphids inwheat Triticum aestivum L (Harwood et al 2004Nienstedt and Poehling 2004) and minute pirate bugsare important predators of soybean aphid (Costama-gna and Landis 2007)

The Chualar and Spreckels trials illustrate instancesinwhichpredatorsother thansyrphidsmaycontributesigniTHORNcantly to the suppression of aphids in organicromaine However unlike the Chualar site whereaverage syrphid densities were among the highest ofall trials 2 wk before harvest syrphid densities werevery low at Spreckels The Spreckels trial also wasunusual in that the aphid infestation in the untreatedplots at harvest approached unmarketable levels Con-sidering the paucity of syrphid larvae the abundance

Table 6 Multiple linear regression analysis and parameter estimates SE for relationships between aphid (y) and syrphid larva (x)densities and between aphid (y) and nonsyrphid predator (z) densities

Factors and sources ofvariation

dfF statistic P value

Numerator Denominator

Trial 4 864 561 0001Syrphid Larvae 1 542 1792 00001Nonsyrphid predators 1 493 000 0982Syrphid larvae trial 4 567 142 0226Nonsyrphid predators trial 4 512 260 0035Sampling wk 4 467 1015 00001

SigniTHORNcance of slope parameter

Parameter estimates DF t value P value

Syrphid larvay 288(009) 0091(002)x 542 423 00001Nonsyrphid predator trialChualar y 298(021) 0302(011)z 505 278 0006Hollister y 330(021) 0080(009)z 486 090 0371Hartnell y 218(021) 0018(017)z 487 011 0914Watsonville y 349(022) 0018(012)z 510 015 0882Spreckels y 288(021) 0179(013)z 546 134 0181

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1531

of minute pirate bugs and the poor biological controlof aphids achieved at this site the Spreckels trial mayoffer an example of interguild predation interferingwith suppression of aphids by syrphids in organicromaine Interguild predation interferes with conser-vation biological control in some instances (Rosen-heim 1998) and further study is required to determinewhether the presence of certain nonsyrphid predatorsreduces the efTHORNcacy of syrphids in suppressing aphidsin organic romaine

It is worth noting that nonsyrphid predators werestatistically higher in romaine treated with Entrust inthe Hollister and Watsonville trials but that Entrust-treated romaine was still unmarketable in these trialsbecause of aphid infestation Numbers of aphid mum-mies also were higher in romaine treated with Entrustin two trials presumably because the densities ofaphids were much higher in romaine treated withEntrust

Given the abundance of syrphid larvae and thatsyrphid larvae are the only predator consistently re-covered from organically grown romaine we arguethat syrphid larvae are primarily responsible for sup-pressing N ribisnigri and other aphids to below eco-nomic levels

Acknowledgments

We thank Robert Bugg (University of California Sustain-able Agriculture Research and Development Program) andDiana Henderson Reisa Bigelow Franklin Dlott Cathy Carl-son Lourdes Rodriguez and Jason Hoeksema for assistancewith this research Sanford Eigenbrode and three anonymousreviewers helped improve the original manuscript We alsothank the farm managers who collaborated with on-farmtrials This research was supported in part by the UnitedStates Environmental Protection Agency Region 9

References Cited

Arthurs S P L A Lacey and E R Miliczky 2007 Eval-uation of codling moth granulosis virus and spinosad forcodling moth control and impact on non-target species inpear orchards Biol Control 41 99ETH109

California Farm Bureau 2006 California Farm Bureau(wwwcfbfcominfoagfacftscfm)

Chaney W E 1999 Lettuce aphid update MontereyCounty Crop Notes University of California CooperativeExtension Salinas CA

Colfer R 2004 Using habitat management to improvebiological control on commercial organic farms in Cal-

ifornia pp 55ETH62 In M S Hoddle [ed] Fourth Cal-ifornia Conference on Biological Control 13ETH15 July2004 Berkeley CA University of California PressBerkeley CA

Contreras J V Quinto V Abellan J Fernandez E Grava-los C Moros and P Bielza 2006 Effects of naturalinsecticides on Frankliniella occidentalis and Orius sppBull OILBSROP 29 331ETH336

Costamagna A C andDA Landis 2007 Quantifying pre-dation on soybean aphid through direct THORNeld observa-tions Biol Control 42 16ETH24

Eizen G W P J Elzen and E G King 1998 Laboratorytoxicity of insecticide residues to Orius insidiosus Geo-coris punctipes Hippodamia convergens and Chrysoperlacarneapp 1235ETH1238 InP Dugger and D Richter [eds]1998 Proceedings of the Beltwide Cotton Conference5ETH9 January 1998 San Diego CA

Harwood J D K D Sunderland and WOC Symondson2004 Prey selection by linyphiid spiders moleculartracking of the effects of alternative prey on rates of aphidconsumption in the THORNeld Mol Ecol 13 3549ETH3560

Klonsky K and K Richter 2007 Statistical review of Cal-ifornia Organic Agriculture 2000ETH2005 Agricultural Is-sues Center University of California Davis CA (aicuc-davisedu)

Liu Y-B 2004 Distribution and population developmentof Nasonovia ribisnigri in iceberg lettuce J Econ Ento-mol 97 883ETH890

MacKenzie J R and R S Vernon 1988 Sampling for thedistribution of lettuce aphid Nasonovia ribisnigri (Ho-moptera Aphididae) in THORNelds and within heads J En-tomol Soc Br Columbia 85 10ETH14

Miles M 2006 The effects of spinosad on beneTHORNcial insectsand mites used in integrated pest management systems ingreenhouses Bull OILSROP 29 53ETH59 [NASS] Na-tional Agricultural Statistics Service 2006 National Ag-ricultural Statistics Service (wwwnassusdagov)

Nienstedt K M and H-M Poehling 2004 Invertebratepredation of 15N-marked prey in semi-THORNeld wheat enclo-sures Entomol Exp Appl 112 191ETH200

Rosenheim J A 1998 High order predators and the regu-lation of insect herbivore populations Annu Rev Ento-mol 43 421ETH47

SAS Insitute 2002ndash2003 SAS version 91 SAS InstituteCary NC

Smith H A and W E Chaney 2007 A survey of syrphidpredators ofNasonovia ribisnigri in organic lettuce on theCentral Coast of California J Econ Entomol 100 39ETH48

Studebaker G E and T J Kring 2003 Effects of insecti-cides on Orius insidiosus measured by THORNeld greenhouseand Petri dish bioassays Fla Entomol 86 178ETH185

Received 21 August 2007 accepted 27 April 2008

1532 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

dae) brown lacewings (Hemerobiidae) rove beetles(Staphylinidae) and big-eyed bugs (Geocoris sp)The complex of nonsyrphid predators was different ateach site (Table 3)ParasitismAphid mummies were found only in the

Hartnell and Watsonville trials Average whole plantdensities of aphid mummies were higher in romainetreated with Entrust (143 022) than in untreatedromaine (008 022) at the Hartnell trial (F 1871df 1 6 P 0005) At the Watsonville trial averagewhole plant densities of aphid mummies were signif-icantly higher in romaine treated with Entrust only inweek 5 (Entrust 325 049 untreated 020 049 F 2113 df 1 6 P 0004)Aphid Density Relationships with Syrphid Larvaeand Nonsyrphid Predators Regression analysis re-vealed a signiTHORNcant negative relationship betweenaphid and syrphid densities in all trials but a signiTHORNcantnegativerelationshipbetweennonsyrphidpredatorsandaphiddensities in theChualar trialonly(Table6)Of thefourothertrialswefoundnorelationshipbetweenaphidand nonsyrphid predator densities

Discussion

On CaliforniaOtildes Central Coast several differenttypes of predators are found in organically grownromaine that is infested with N ribisnigri and otheraphids The data presented here and previously(Smith and Chaney 2007) indicate that syrphid larvaeare the most abundant predators often making up85 of the predators collected from infested ro-maine and that they can average up to nine larvae perromaine head during the crucial weeks immediatelybefore the harvest of the crop In addition syrphidlarvae are the only predator group consistently foundin all infested romaine THORNelds Other predators are lessabundant and they are recovered in some THORNelds butnot others Regression analysis presented here sug-gests that increases in syrphid larvae help explainreductions in aphid numbers and that there is no

relationship between nonsyrphid predators and aphiddensities at most sites

We present data that demonstrates that aphid den-sities were signiTHORNcantly higher in plots where syrphidswere suppressed Romaine in which syrphids weresuppressed was unmarketable because of aphid infes-tation whereas untreated romaine was marketableOur results suggest that syrphids play a major role inthe conservation biological control ofN ribisnigri andother aphids in organically grown romaine on theCentral Coast of California

We found that although Entrust does not affectoviposition by syrphid females it does suppress non-syrphid predators in some instances Because Entrustapparently reduced the densities of nonsyrphid pred-ators as well as syrphid larvae at the Chualar andSpreckels trials it was an imperfect tool for demon-strating the effect of selectively removing syrphid lar-vae from organically grown romaine Given thatminute pirate bugs composed 30 of the predators atthe Spreckels trial in untreated romaine and only 12where Entrust was applied it is possible that suppres-sion of minute pirate bugs contributed to the releaseof aphids from predation in the Entrust treatmentDwarf spiders and minute pirate bugs in the Chualartrial also may have contributed to aphid suppressionDwarf spiders are important predators of aphids inwheat Triticum aestivum L (Harwood et al 2004Nienstedt and Poehling 2004) and minute pirate bugsare important predators of soybean aphid (Costama-gna and Landis 2007)

The Chualar and Spreckels trials illustrate instancesinwhichpredatorsother thansyrphidsmaycontributesigniTHORNcantly to the suppression of aphids in organicromaine However unlike the Chualar site whereaverage syrphid densities were among the highest ofall trials 2 wk before harvest syrphid densities werevery low at Spreckels The Spreckels trial also wasunusual in that the aphid infestation in the untreatedplots at harvest approached unmarketable levels Con-sidering the paucity of syrphid larvae the abundance

Table 6 Multiple linear regression analysis and parameter estimates SE for relationships between aphid (y) and syrphid larva (x)densities and between aphid (y) and nonsyrphid predator (z) densities

Factors and sources ofvariation

dfF statistic P value

Numerator Denominator

Trial 4 864 561 0001Syrphid Larvae 1 542 1792 00001Nonsyrphid predators 1 493 000 0982Syrphid larvae trial 4 567 142 0226Nonsyrphid predators trial 4 512 260 0035Sampling wk 4 467 1015 00001

SigniTHORNcance of slope parameter

Parameter estimates DF t value P value

Syrphid larvay 288(009) 0091(002)x 542 423 00001Nonsyrphid predator trialChualar y 298(021) 0302(011)z 505 278 0006Hollister y 330(021) 0080(009)z 486 090 0371Hartnell y 218(021) 0018(017)z 487 011 0914Watsonville y 349(022) 0018(012)z 510 015 0882Spreckels y 288(021) 0179(013)z 546 134 0181

October 2008 SMITH ET AL SYRPHIDS IN ORGANIC LETTUCE 1531

of minute pirate bugs and the poor biological controlof aphids achieved at this site the Spreckels trial mayoffer an example of interguild predation interferingwith suppression of aphids by syrphids in organicromaine Interguild predation interferes with conser-vation biological control in some instances (Rosen-heim 1998) and further study is required to determinewhether the presence of certain nonsyrphid predatorsreduces the efTHORNcacy of syrphids in suppressing aphidsin organic romaine

It is worth noting that nonsyrphid predators werestatistically higher in romaine treated with Entrust inthe Hollister and Watsonville trials but that Entrust-treated romaine was still unmarketable in these trialsbecause of aphid infestation Numbers of aphid mum-mies also were higher in romaine treated with Entrustin two trials presumably because the densities ofaphids were much higher in romaine treated withEntrust

Given the abundance of syrphid larvae and thatsyrphid larvae are the only predator consistently re-covered from organically grown romaine we arguethat syrphid larvae are primarily responsible for sup-pressing N ribisnigri and other aphids to below eco-nomic levels

Acknowledgments

We thank Robert Bugg (University of California Sustain-able Agriculture Research and Development Program) andDiana Henderson Reisa Bigelow Franklin Dlott Cathy Carl-son Lourdes Rodriguez and Jason Hoeksema for assistancewith this research Sanford Eigenbrode and three anonymousreviewers helped improve the original manuscript We alsothank the farm managers who collaborated with on-farmtrials This research was supported in part by the UnitedStates Environmental Protection Agency Region 9

References Cited

Arthurs S P L A Lacey and E R Miliczky 2007 Eval-uation of codling moth granulosis virus and spinosad forcodling moth control and impact on non-target species inpear orchards Biol Control 41 99ETH109

California Farm Bureau 2006 California Farm Bureau(wwwcfbfcominfoagfacftscfm)

Chaney W E 1999 Lettuce aphid update MontereyCounty Crop Notes University of California CooperativeExtension Salinas CA

Colfer R 2004 Using habitat management to improvebiological control on commercial organic farms in Cal-

ifornia pp 55ETH62 In M S Hoddle [ed] Fourth Cal-ifornia Conference on Biological Control 13ETH15 July2004 Berkeley CA University of California PressBerkeley CA

Contreras J V Quinto V Abellan J Fernandez E Grava-los C Moros and P Bielza 2006 Effects of naturalinsecticides on Frankliniella occidentalis and Orius sppBull OILBSROP 29 331ETH336

Costamagna A C andDA Landis 2007 Quantifying pre-dation on soybean aphid through direct THORNeld observa-tions Biol Control 42 16ETH24

Eizen G W P J Elzen and E G King 1998 Laboratorytoxicity of insecticide residues to Orius insidiosus Geo-coris punctipes Hippodamia convergens and Chrysoperlacarneapp 1235ETH1238 InP Dugger and D Richter [eds]1998 Proceedings of the Beltwide Cotton Conference5ETH9 January 1998 San Diego CA

Harwood J D K D Sunderland and WOC Symondson2004 Prey selection by linyphiid spiders moleculartracking of the effects of alternative prey on rates of aphidconsumption in the THORNeld Mol Ecol 13 3549ETH3560

Klonsky K and K Richter 2007 Statistical review of Cal-ifornia Organic Agriculture 2000ETH2005 Agricultural Is-sues Center University of California Davis CA (aicuc-davisedu)

Liu Y-B 2004 Distribution and population developmentof Nasonovia ribisnigri in iceberg lettuce J Econ Ento-mol 97 883ETH890

MacKenzie J R and R S Vernon 1988 Sampling for thedistribution of lettuce aphid Nasonovia ribisnigri (Ho-moptera Aphididae) in THORNelds and within heads J En-tomol Soc Br Columbia 85 10ETH14

Miles M 2006 The effects of spinosad on beneTHORNcial insectsand mites used in integrated pest management systems ingreenhouses Bull OILSROP 29 53ETH59 [NASS] Na-tional Agricultural Statistics Service 2006 National Ag-ricultural Statistics Service (wwwnassusdagov)

Nienstedt K M and H-M Poehling 2004 Invertebratepredation of 15N-marked prey in semi-THORNeld wheat enclo-sures Entomol Exp Appl 112 191ETH200

Rosenheim J A 1998 High order predators and the regu-lation of insect herbivore populations Annu Rev Ento-mol 43 421ETH47

SAS Insitute 2002ndash2003 SAS version 91 SAS InstituteCary NC

Smith H A and W E Chaney 2007 A survey of syrphidpredators ofNasonovia ribisnigri in organic lettuce on theCentral Coast of California J Econ Entomol 100 39ETH48

Studebaker G E and T J Kring 2003 Effects of insecti-cides on Orius insidiosus measured by THORNeld greenhouseand Petri dish bioassays Fla Entomol 86 178ETH185

Received 21 August 2007 accepted 27 April 2008

1532 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5

of minute pirate bugs and the poor biological controlof aphids achieved at this site the Spreckels trial mayoffer an example of interguild predation interferingwith suppression of aphids by syrphids in organicromaine Interguild predation interferes with conser-vation biological control in some instances (Rosen-heim 1998) and further study is required to determinewhether the presence of certain nonsyrphid predatorsreduces the efTHORNcacy of syrphids in suppressing aphidsin organic romaine

It is worth noting that nonsyrphid predators werestatistically higher in romaine treated with Entrust inthe Hollister and Watsonville trials but that Entrust-treated romaine was still unmarketable in these trialsbecause of aphid infestation Numbers of aphid mum-mies also were higher in romaine treated with Entrustin two trials presumably because the densities ofaphids were much higher in romaine treated withEntrust

Given the abundance of syrphid larvae and thatsyrphid larvae are the only predator consistently re-covered from organically grown romaine we arguethat syrphid larvae are primarily responsible for sup-pressing N ribisnigri and other aphids to below eco-nomic levels

Acknowledgments

We thank Robert Bugg (University of California Sustain-able Agriculture Research and Development Program) andDiana Henderson Reisa Bigelow Franklin Dlott Cathy Carl-son Lourdes Rodriguez and Jason Hoeksema for assistancewith this research Sanford Eigenbrode and three anonymousreviewers helped improve the original manuscript We alsothank the farm managers who collaborated with on-farmtrials This research was supported in part by the UnitedStates Environmental Protection Agency Region 9

References Cited

Arthurs S P L A Lacey and E R Miliczky 2007 Eval-uation of codling moth granulosis virus and spinosad forcodling moth control and impact on non-target species inpear orchards Biol Control 41 99ETH109

California Farm Bureau 2006 California Farm Bureau(wwwcfbfcominfoagfacftscfm)

Chaney W E 1999 Lettuce aphid update MontereyCounty Crop Notes University of California CooperativeExtension Salinas CA

Colfer R 2004 Using habitat management to improvebiological control on commercial organic farms in Cal-

ifornia pp 55ETH62 In M S Hoddle [ed] Fourth Cal-ifornia Conference on Biological Control 13ETH15 July2004 Berkeley CA University of California PressBerkeley CA

Contreras J V Quinto V Abellan J Fernandez E Grava-los C Moros and P Bielza 2006 Effects of naturalinsecticides on Frankliniella occidentalis and Orius sppBull OILBSROP 29 331ETH336

Costamagna A C andDA Landis 2007 Quantifying pre-dation on soybean aphid through direct THORNeld observa-tions Biol Control 42 16ETH24

Eizen G W P J Elzen and E G King 1998 Laboratorytoxicity of insecticide residues to Orius insidiosus Geo-coris punctipes Hippodamia convergens and Chrysoperlacarneapp 1235ETH1238 InP Dugger and D Richter [eds]1998 Proceedings of the Beltwide Cotton Conference5ETH9 January 1998 San Diego CA

Harwood J D K D Sunderland and WOC Symondson2004 Prey selection by linyphiid spiders moleculartracking of the effects of alternative prey on rates of aphidconsumption in the THORNeld Mol Ecol 13 3549ETH3560

Klonsky K and K Richter 2007 Statistical review of Cal-ifornia Organic Agriculture 2000ETH2005 Agricultural Is-sues Center University of California Davis CA (aicuc-davisedu)

Liu Y-B 2004 Distribution and population developmentof Nasonovia ribisnigri in iceberg lettuce J Econ Ento-mol 97 883ETH890

MacKenzie J R and R S Vernon 1988 Sampling for thedistribution of lettuce aphid Nasonovia ribisnigri (Ho-moptera Aphididae) in THORNelds and within heads J En-tomol Soc Br Columbia 85 10ETH14

Miles M 2006 The effects of spinosad on beneTHORNcial insectsand mites used in integrated pest management systems ingreenhouses Bull OILSROP 29 53ETH59 [NASS] Na-tional Agricultural Statistics Service 2006 National Ag-ricultural Statistics Service (wwwnassusdagov)

Nienstedt K M and H-M Poehling 2004 Invertebratepredation of 15N-marked prey in semi-THORNeld wheat enclo-sures Entomol Exp Appl 112 191ETH200

Rosenheim J A 1998 High order predators and the regu-lation of insect herbivore populations Annu Rev Ento-mol 43 421ETH47

SAS Insitute 2002ndash2003 SAS version 91 SAS InstituteCary NC

Smith H A and W E Chaney 2007 A survey of syrphidpredators ofNasonovia ribisnigri in organic lettuce on theCentral Coast of California J Econ Entomol 100 39ETH48

Studebaker G E and T J Kring 2003 Effects of insecti-cides on Orius insidiosus measured by THORNeld greenhouseand Petri dish bioassays Fla Entomol 86 178ETH185

Received 21 August 2007 accepted 27 April 2008

1532 JOURNAL OF ECONOMIC ENTOMOLOGY Vol 101 no 5