BIOLOGICALCONTROL

Melon Aphid (Homoptera Aphididae) Control by InundativeConvergent Lady Beetle (Coleoptera Coccinellidae)

Release on Chrysanthemum

STEVE H DREISTADTI AND MARY LOUISE FLlNT12

EnvironEntomo25(3)688-697(1996)ABSTRACT Convergent lady beetles Hippodamia convergcns Cuerin-Meneville collectedfrom aggregations in California were released for control of melon aphids Aphis gossypiiClover infesting potted chrysanthemum Dendranthema grandiflora (Tzvelev) Hurricaneoutdoors Most beetles dispersed from plants within 1-3 d after a single release during May1993 and July and September 1993 and 1994 Most dispersing beetles left the study areaalthough some moved within release plots to adjoining plants or dispersed to nearby plotswithin =40 m Aggregation-collected beetles that were allowed to drink and fly in a screentent for 7-10 d before release and insectary-reared beetles dispersed more slowly than aggre-gation-collected beetles that were stored at 4-lOoC until the day of release Although beetlesdispersed within days releases significantly reduced aphid densities during all trials A singlerelease of 34-42 adult H convergem per pot provided 25-84 aphid control calculatedaccording to the Abbott metilOd Each beetle consumed =25-170 melon aphids per dayTherewas a density-dependent functional response in predation beetles consumed more aphidswhen released on plants with higher aphid densities Release of commercially available con-vergent lady beetles can provide augmentative control of relatively high aphid densities onsmall potted plants

KEY WORDS Hippodamia convergellS Aphis gossypii Dendranthema grandiflora inun-dative control

THE CONVERGENTLADYbeetle Hippodamia con-vergens Guerin-Meneville is widely released foraphid control Large aggregations in the Sierra Ne-vada allow convergent lady beetles to be readilycollected and packaged for sale in nurseries gar-den supply catalogs and other outlets across theUnited States No data are available but our com-mercial supplier reportedly sells 20000 liters(5000 gal) of convergent lady beetles each yearand estimates that these sales are 10 of theCalifornia market Because 1 gal contains 70000-80000 beetles several billion convergent lady bee-tles are apparently collected and sold each year

The University of California has histOrically rec-ommended against releasing aggregation-collectedconvergent lady beetles because releases were be-lieved ineffective (Moore and Koehler 1981Hagen 1982) The recommendation against releas-es was based mostly on observations that H con-vergens collected from aggregations dispersed soonafter being released Davidson (1919 1924) re-leased marked beetles in barley Hordeum vulgareL and wheat Triticum sp and found that only10 of beetles remained after 1 wk Eddy (1939)

lIPMEducationandPublicationsStatewideIPMProjectUni-versityof CaliforniaDavisCA95616-8620

2Departmentof EntomologyUniversityof CaliforniaDavisCA95616

reported that most H convergens quickly dis-persed after release in sugar cane Saccharum of-

ficinarum L Packard and Campbell (1926) foundthat a few of the H convergens released in alfalfaMedicago sativa L remained after 12 d and thatthe postrelease density of pea aphids Acyrthosi-phon pisum (Harris) apparently declined Cooke(1963) credited H convergens release for someaphid reduction in alfalfa but like Fenton andDahms (1951) and Hatch and Tanasse (1948)Cooke concluded that releases would not be eco-nomical in comparison with insecticides Starks etal (1975) found that 72 of H convergens dis-persed immediately after release in sorghum Sor-ghum vulgare Persoon in Kansas although dis-persal was slowed by providing a shelter of wetburlap releases did not significantly reduce aphiddensities

Except for Starks et al (1975) none of theseauthors provide any data on aphid densities andnone report using nonrelease control plots in theirexperimental designs Most early researchers wereinvestigating inoculative control and because mostbeetles soon dispersed without reproducing theseworkers concluded that releases would be ineffec-tive At least some workers did not consider it pos-sible that the beetles they released could provideinundative control For example Packard and

0046-225X960688-0697$02000copy 1996EntomologicalSocietyofAmerica

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 689

Campbell (1926) observed a decrease in peaaphids but did not credit this to the release be-cause it was too soon after liberation for progenyof the introduced beetles to have become effec-tive Recent research focused on dispersal and ef-fects of temperature wind and nutrition on dis-persal of H convergens (Hagen 1962 Rankin andRankin 1980 Davis and Kirkland 1982)

Despite long-standing opinions that H conver-gens releases are ineffective apparently onlyRaupp et aI (1994) and Flint et al (1995) haveconducted replicated controlled experiments onaphid density changes from release of H conver-gens Raupp et al (1994) released 10 H conver-gens per plant on potted firethorn Pyracantha le-landii infested with spirea aphid Aphis spiraecolaPatch and a woolly aphid (Eriosoma sp) in thegreenhouse Release controlled only spirea aphidsand only if phmts were caged An uncaged fieldrelease of1500 beetles by Raupp et al (1994) ap-parently reduced Capitophorus elaeagni (delCuerico) aphids on an Elaeagnus pungens Thun-berg hedge

Flint et aI (1995) reported significant reductionsin numbers of melon aphid Aphis gossypii Cloveron chrysanthemum Dendranthema grandifloraTzvelev and rose aphid Macrosiphum rosae (L)on rose Rosa sp when beetles were released out-doors on uncaged potted plants Here we reporton our investigation of whether dispersal of aggre-gation-collected H convergens was reduced by fly-ing and feeding beetles before releasing them Wecompared dispersal of insectary-reared beetleswith dispersal of beetles from aggregations and de-termined whetller inundative release of convergentlady beetles controls aphids on potted plants out-doors

Materials and Methods

We released marked adult convergent lady bee-tles outdoors on potted chrysanthemums infestedwith melon aphids We conducted 5 trials eachtrial of an experiment was conducted once duringMay July and September 1993 and July and Sep-tember 1994 Except for the insectary-reared bee-tles discussed below beetles were from Unique In-sect Control Citms Heights CA This commercialsupplier reports that beetles were from aggrega-tions at an altitude of 600 m in Placer CountyCalifornia and that beetles were collected 52 wkbefore each trial in July and September Beetlesreleased in May reportedly were collected in lateFebruary 1993 and stored at 4-100C

Treatment Effects on Beetle Dispersal Wepreconditioned (treated) beetles before releaseand conducted mark-release-recapture studies tocompare dispersal tendencies among treatmentsAll 5 trials included the following 2 treatmentsconducted with aggregation-collected lady beetles(1) beetles were allowed to fly for 1 wk (during1993) or 10 d (1994) in a screen-wall tent (3 by 3

by 2 m) outdoors sprayed daily with diluted honeyand marked before release (flown) or (2) beetleswere removed from cold storage (4-100C) 8 hbefore release confined witll honey and watermarked before release and sprayed with soda popimmediately after release (stored) Soda pop wasapplied because beetle distributors commonly rec-ommend application of this or other sweet liquidsto reduce beetle dispersal Third treatments wereadded to the trials during May 1993 and Septem-ber 1994 as described below

The May 1993 trial included a 3rd treatment ofbeetles produced in the insectary (reared) to de-termine whether rearing beetles influenced dis-persal in comparison with aggregation-collectedbeetles Reared H convergens adults were the Ist-generation offspring from adults collected in alfalfain Davis (Yolo County) in April Field-collectedadults and their progeny were held in petri dishesat 20degC and continuous light They were providedhoney water and ad lib an assortment of field-collected aphids mostly Aphis fabae Scopoli Aphisgossypii and green peach aphid Myzus persicae(Sulzer) After eclosion adults were held and fedthe same diet as larvae for 1-2 wk before release

The 3rd treatment during September 1994 wasdesigned to assess whether marking influencedbeetle dispersal Aggregation-collected beetleswere released but not marked before release (un-marked) They were otherwise handled the sameas stored beetles

To distinguish flown from stored beetles wemarked beetles before release by spraying 1 treat-ment with white Ace Instant Drying InteriorEx-terior Lacquer (Ace Hardware Oak Brook IL)Dark (black or blue) Ace Instant Drying InteriorExterior Lacquer was applied to beetles in the 2ndtreatment We also used aluminum Zynolyte SprayLacquer (Zynolyte Carson CA) during May whena 3rd middottreatment marking was needed

To apply lacquer we confined beetles from eachtreatment in a separate cardboard box (30 by 22by 5 cm) with window screen on 1 side andsprayed lacquer into the box through the screenonto beetles for 2 s We aired the box for 30 sto dry the lacquer and disperse solvent shook bee-tles down onto the bottom of the box allowed bee-tles to right themselves so that their elytra wereexposed and resprayed them We repeated thisspray-and-shake procedure 4 times until mostbeetles appeared to be marked

To assure ourselves of the effectiveness of mark-ing we retained and inspected beetles from treat-ments before release during July 1994 and Sep-tember 1993 and 1994 We collected 5-8 samplesof 17 (during 1993) or 40 (1994) beetles each fromevery treatment during each trial and examinedbeetles for marking using a dissecting binocularmicroscope Except for the unmarked treatmentduring September 1994 964 (SD = 46) of thebeetles recovered were marked We do not reportunmarked beetles recovered in the numbers pre-

690 ENVIRONMENTALENTOMOLOGY Vol 25 no 3

XX XX XX XX XXX 000 X 000 X 000 X 000 X 000X 000 X 000 X 000 X 000 X 000

XX XX XX XX XX

1993 10 m

xx xx xx xx xxX 000 X 000 X 000 X 000 X 000X 000 X 000 X 000 X 000 X 000

xx xx xx xx xx

XXXxoooo 0xoooo 0xxx

X X XxoooooxoooooX X X

X X Xxoooooxooooo

X X X

X X X X X X X X Xxooooo xooooo xoooooxooooo xooooo xooooo

X X X X X X X X XPlant Pots 1994 3m

X Rose x x x xxx xxxo Chrysanthemum Xooooo Xooooo XoooooXooooo xooooo xooooo

X X X X X X X X XFig 1 Convergent lady beetle experimental design Release and control plots were a randomized block 2 X 5

plots during 1993 (above within rectangle) and 3 x 3 plots during 1994 (below) except as described in Materialsand Methods Each 4-liter pot (0) contained 4 chrysanthemum plants with 6 (during 1993) or 10 (1994) pots perplot Each plot was sheltered on 3 sides by 6 (1993)or 8 (1994) aphid-free 1-m-tall rose Rosa sp Tropicana plantsin 20-liter pots (x) Beetle dispersal out of plots was evaluated by collecting all beetles from 1 or 2 previouslyunsampled pots in each release plot at intervals of 12 24 36 48 and 69 h after release Beetle movement withinplots was evaluated at each interval by resampling and removing beetles from all previously sampled pots

sented here except during September 1994 whenall unmarked beetles we recovered were presumedto be from the unmarked treatment beetles we re-leased

Beetle Releases We conducted all experimentsoutdoors at the University of California Davis Ar-boretum nursery in a gravel-surfaced area (130 by80 m) The area was free of plants except for thosein our study Each 4-liter pot contained 4 chrysan-themum plants We used 6 (1993) or 10 (1994)chrysanthemum pots (24 or 40 plants) in each plotsurrounded on 3 sides by potted roses Roses wereaphid-free and were used to shade the chrysanthe-mums which had been produced in a whitewashedgreenhouse Chrysanthemum pots did not touchwithin plots there was a gap (3-8 cm) between thecanopy of each pot Distance from plot edge to the

nearest other plots was 10 m (during 1993) or 3 In(1994) (Fig 1)

We counted the number of beetles to be re-leased into each pot by aspirating them into a vialselecting only apparently healthy (active) beetlesthat appeared marked We released beetles hyknocking them from vials onto potting media at thebase of the plants We released the stored beetlessprayed them with 10 ml of soda pop per potthen released the other treatment beetles in thatsame pot (during 1993) before moving to releasebeetles in the next pot Foliage and media werewet from watering plants immediately before re-lease

We released 34-42 beetles per pot (see h = 0in Results for exact numbers) once at dusk on 18May 20 July and 28 September 1993 and on 26

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 691

July and 27 September 1994 During 1993 we re-leased equal numbers of flown and stored H con-vergens in each pot During May when we alsonleased reared beetles 13 of the beetles releasedin each pot were of each treatment During 1994each release plot received only beetles from 1treatment either all flown or all stored beetles dur-ing July or all flown stored or unmarked beetlesduring September Release plots were replicated 6times during 1993 and 3 times for each treatmentduring 1994 as illustrated in Fig 1 Nonrelease(control) plots were replicated 4 (1993) or 3 (1994)times except that the May 1993 trial included only1 plot of control pots A randomized block of 2 X5 plots was used during 1993 3 X 3 during July1994 and 3 X 4 during September 1994 the for-mer 2 designs are illustrated in Fig 1 Before re-lease we controlled ants around all plants by de-ploying Grants Kills Ants (003 arsenic) (GrantLaboratories San Leandro CAl ant stakes

Beetle Dispersal At 12 24 36 48 and 69 hafter release we evaluated dispersal by aspiratingall beetles from 1 (during 1993) or 2 (1994) pre-viously unsampled pots in each plot Two excep-tions were as follows (1) during September 1993we collected beetles from 2 previously unsampledpots per plot at 69 h and (2) during July 1993 wecollected beetles at 12 24 and 48 h and beetleswere collected from 4 previously unsampled potsper plot at 48 h

Nondispersing beetles (none) were those recov-ered on previously unsampled pots and marked thesame color as the beetles released in that plot Wemeasured dispersal within plots at these same in-tervals by rcsampling and removing beetles fromall previously sampled pots for example at 36 hafter release during 1994 we removed all beetlesfrom 2 previously unsampled pots in each plot andremoved beetles that had moved onto pots sam-pled 12 and 24 h earlier To measure dispersalamong plots we recorded whether the beetles werecovered were marked with the same color as thebeetles released into that plot

Aphid Control We grew chrysanthemums inthe greenhouse under a photoperiod of 168 (LD)h at 20-30degC We planted 4 cuttings in each 4-literpot grew them for 8-10 wk then moved plantsoutdoors to the study area the day of release Weinoculated plants vith A gossypii 4--5wk after pot-ting and excluded other insects by screening yel-low sticky traps or removal of infested leaves

At ~36 h before release (to) we sampled aphidson 3 or 4 pots per plot during 1993 and on 4 or 6pots per plot during 1994 We resampled 2 potsper plot at 36 h after release during 1994 and re-sampled 2-4 pots per plot 72h after release dur-ing 1993 and 1994 The release plants sampled foraphids at 36 h were those from which we had justremoved beetles for the dispersal assessment de-scribed previously Release pots resampled 72 hafter release were those from which beetles were1st removed at 48 and 69 h after release except

that during September 1994 aphids were not sam-pled on pots from which beetles were first re-moved at 48 h

We counted all aphids on 1 leaf per plant (4leaves per pot) Each sample leaf was from abouthalfWayup the main stem from the soil line be-cause middle-canopy leaves consistently provide anaverage measure of aphid density per plant (Vehrset al 1992) During July 1994 we counted aphidson each 12 (on either side of the midvein) of 1middle canopy leaf on 32 plants and used a pairedt-test to determine whether there was significantdifference (t = 019 P = 085) in aphid densitybetween halves of the same leaf For the remainingplants during July and all plants during September1994 we counted aphids on only 12 of 1 leaf perplant and multiplied this value by 2 to obtainwhole-leaf estimates

In May 1993 we randomly allocated pots to re-lease or nonrelease plots During July and Septem-ber of both years before release we sampledaphids on 2 times the number of pots neededWe averaged the aphid density on the 4 plants perpot rank-ordered pots according to aphid densitydiscarded pots with the highest and lowest aphiddensities and allocated pots so that the averageprerelease aphid density was not significantly dif-ferent among plots We randomized pot locationwithin plots and randomly assigned plots to con-trols or releases

We counted the number of leaves and measuredthe height of 1 plant per pot We multiplied thenumber of leaves per plant by 4 plants per pot toestimate leaves per pot (LPP) We estimated dailyaphid consumption per beetle (AB)

AB = [ARto - ARtn) - (ACto - ACtn)] X LPP(Bto + Btn)2

where AR and AC are aphids per leaf on releaseand control plants respectively and B is beetlesper pot at release time (to) or 36 or 72 h afterrelease (tn) We included (AClo - ACtn) in theformula to adjust estimates of aphid consumptionThis adjustment was necessary because even in theabsence of lady beetles aphids declined during 4of 5 trials probably because of hot dry conditionsin the field or because some other aphid predatorscolonized plants after they were moved outdoorsor both

To determine whether there was a density-de-pendent functional response in predation we usedleast squares linear regression analyses (PROCREG SAS Institute 1988) to examine the relation-ship between aphid density at the time of release(ARto) (the independent variable) and daily aphidconsumption per beetle (AB) During 1993 we de-termined AB for all treatments combined duringeach of 3 trials (May July and September) be-cause all treatments were released into the sameplots Treatments were released in different plotsduring 1994 so we were able to calculate AB sep-

692 ENVIHONMENTAL ENTOMOLOGY Vol 25 no 3

Fig 2 Mean + SEM number of adult lady beetlesrecoveredafter a single release on chrysanlliemumsin-festedwillimelonaphidsfor trialsconductedin May(A)July (B) and September (e) 1993 Beetles were insec-tary-reared (reared) collected from aggregationsandcold-storeduntil the day of release (stored)or collectedfrom aggregationsand conditioned(flown)

beetles we recovered were in the same plots wherethey were released Because we also resampledplants from which we had previously removed allbeetles we found that when beetles did dispersethey usually left our study site (Fig 4 out cate-gory) rather ilian moving to nearby plants Most ofthe beetles iliat we did recover were on plants inthe pots where they were released (Fig 4 dis-persal category none) Fifteen percent or less ofthe beetles dispersed to oilier plants in ilieir re-lease plot (Fig 4 within category) Five percentor less of beetles were recovered in nearby plots(Fig 4 other category) Flown beetles weremore likely to disperse locally in comparison withstored beetles about twice as many flown beetleswere recovered on other plants (within category)in the release plot or in nearby plots (oilier cat-egory) (Fig 4) in comparison witll stored beetles

Marking beetles wiili lacquer to distinguishtreatments had no apparent effect on aphid con-

B

69

c

A

Reared

c=J Flown

- Stored

12 24 36 48

Hours After Release

o

20

10

10

20

o

0 0aCD 20c-ooOJ

~ampl 10-w(f)

-tIX 0

Results

Beetle Dispersal Most adult H convergens re-leased on groups of aphid-infested potted plants(Fig 1) dispersed within 1-3 d (Figs 2 and 3) In2 of 5 trials (Figs 2C and 3A) allowing these ag-gregation-collected beetles to drink and fly (flown)in a screen tent for 7-10 d before release delayeddispersal from plots in comparison with beetlesthat were stored at 4-lOoC (stored) until the dayof release About 20 (SE = 7) and 36 (SE =13) more flown beetles were recovered in com-parison with stored beetles during September 1993(Fig 2C) (t = 228 n = 36 P = 0029) and July1994 (Fig 3A) (t = 260 n = 90 P = 0011)respectively when samples from all postreleasetimes (12-69 h) were combined During July 1993significantly more stored than flown beetles wererecovered during all sample times pooled (Fig 2B)(t = 303 n = 36 P lt 0005)

Insectary-reared beetles dispersed from plotsmore slowly than did aggregation-collected beetles(Fig 2A)About 19 (SE = 10) and 41 (SE =9) more reared beetles were recovered in compar-ison wiili flown (t = 209 n = 30 P = 0046) orstored (t = 389 n = 30 P = 00005) beetlesrespectively when all postrelease samples werepooled during May 1993 (Fig 2A)

Because treatments were marked wiili a differ-ent color and released into separate plots during1994 we were able to determine that most of ilie

arately for each treatment during both 1994 trials(July and September) The result is 8 observationsfor the regression at 72 h There were 5 observa-tions for the regression at 36 h because aphidswere not sampled at that interval during 1993

Percentage control (C) was calculated for eachtreatment according to the Abbott (1925) method

C = [(N - R)N] X 100

where N is the percentage of aphids alive (postre-lease density divided by prerelease density multi-plied by 100) on nonrelease plants and R is thepercentage of aphids alive on release plants at 36or 72 h after release

Statistical Analyses For each treatment wecompared mean aphid densities before and afterreleases and mean numbers of beetles recoveredat all postrelease times combined Beetles in non-release plots were excluded from the data andanalyses Less than 1 of beetles recovered duringeach trial were collected in nonrelease plots Un-less stated otherwise we used t-tests for 2-waycomparisons and analysis of variance (PROCANOVA SAS Institute 1988) for multiple compar-isons Because the beetle dispersal out categoryequals the sum of all other categories subtractedfrom tlle total number of beetles released allmeans were not independent and multiple t-testsinstead of ANOVA were used to compare beetledispersal

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 693

Fig 3 Mean + SEM number of adult convergentlady beetles recovered after a single release on chrysan-themums infested with melon aphids for trials conductedin July (A) and September (B) 1994 Beetles were col-lected from aggregations and were cold-stored until theday of release (stored) cold-stored until release and notlacquered (unmarked) or conditioned (flown)

sumption (see below) or dispersal During Septem-ber 1994 we released cold-stored H convergensthat were both marked (stored) and not marked(unmarked) and there were no obvious differencesin average recovery at individual sample times(Fig 3B) or when all samples were pooled (t =068 n = 120 P = 050) No natural populationsof convergent lady beetles were observed on sur-rounding vegetation in September and presum-ably all unmarked beetles we recovered were thosewe released

Aphid Control Despite relatively rapid dispers-al from plots release of convergent lady beetlessignificantly reduced melon aphid densities on pot-ted chrysanthemums during all trials (Figs 5 and6) Within 3 d after a single release of 34-42 bee-tles per pot H convergens provided 25-84 con-trol as calculated according to the Abbott (1925)method Up to 54 control occurred 15 dafterrelease (Table 1)

Each adult H convergens consumed =25-170melon aphids per day (Fig 7) as estimated by di-viding the average reduction in aphids by the av-erage number of beetles present Samples collect-

A

c=J FloWn

- Storedc=J Unmarked

ILB

III eOther

B

A

c=J Flown

- Stored

Within

e e e f

d

Out

100

80-c~ 60Q)

gt8~ 40CIlQ) 20~Q)

CD-c 0Q)Vllaquol(I) 80Qj00 a~ 600

LiJCI) 40plusmn-IX

20

0None

ed 3 d after release reveal a strong positiveassociation between aphid density per leaf at thetime of release and daily aphid consumption perbeetle (AB) AB was higher on plants with a bigherinitial aphid density (Fig 7) Feeding rate and ini-tial aphid density were not signiRcantly correlatedat 15 d (r2 = 065 F = 56 df = 1 3 P = 01)aphids were sampled at this interval only during1994 and sample size was small

The extent of predation by poundlown beetles in com-parison with stored beetles could not be deter-mined during 1993 because equal numbers of bee-tles from all treatments were released on each potHowever poundlown and stored beetles were releasedinto separate plots during 1994 and both treat-ments significantly reduced aphid densities in com-parison with plots without beetles (Fig 6) Therewere no apparent differences in daily aphid con-sumption per beetle among treatments (poundlown

Dispersal

Fig 4 Mean + SEM dispersal (percentage of re-leased beetles recovered) for flownand stored convergentlady beetles for all sample times pooled over 3 d duringJuly (A) or September (B) 1994 In total 2400 beetleswere released in Julyand 2400 beetles (1200 each flownand stored) were released during September Nondis-persing beetles (none) were recovered in pots where theywere released Beetles that left the study area (out) werenot recovered Some beetles moved from the pot wherethey were released but remained within their release plot(within)The remaining beetles were recovered after dis-persing from release plants to other plots (other) in ourstudy area Within each month means that do not sharea common letter are significantlydifferent (P lt 005 t-tests)

6912 24 36 48Hours After Release

o

40

30

0 20abullQ) 10aCIlQ)Q)Q) 0enLiJ 40CI)Ix 30

20

10

0

694 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

c=J No Beetles - With Beetles A

0250 c==J No Beetles Stored B

- Flown ~ Unmarked200

150

100

50

0Before 15 Days 3 Days

500

-lt1l3 100

300

400

200

bulllt1lC()s1sclaquoWC)+-Ix

B

c

A

80

250200150100

50

- 0~ 50JQ 40a~ 30s~ 20gr 10+I 0

120

Fig 5 Mean + SEM melon aphids per leaf beforeand 3 d after release outdoors of 34-42 adult lady beetlesper pot each pot containing 4 chrysanthemum plants fortrials conducted in May (A) July (B) and September (C)1993 Aphid means in each trial were not significantlydifferent (P gt 001) before release (A) t = 139 n = 1272 (B) t = 031 n = 48 72 (C) t = 008 n = 64 96Means were significantly different (P lt 00001) after re-lease (A) t = 468 n = 20 72 (B) t = 1186 n = 4872 (C) t = 574 n = 6496 Mean plusmn SEM leaves perplant (L) and plant height (H) in centimeters for all treat-ments pooled were (A) L = 154 (04) H = 287 (O(l)(B) L = 158 (04) H = 301 (08) (C) L = 171 (06)H = 297 (11)

40

oBefore Release After Release

Fig 6 Mean + SEM melon aphids per leaf beforeand 15 and 3 d after release outdoors of 40 lown storedor unmarked adult convergent lady beetles per pot eachpot containing 4 chrysanthemum plants for trials con-ducted in July (A) and September (B) 1994 Aphid meansin each trial were not significantly different (P gt 005)before release (A) F = 00 df = 2 213 (B) F = 002df = 3 187 There were significant differences (P lt00001) in aphid densities after release of convergent ladybeetles ANOVA statistics are (A) 15 d F = 195 df =2 69 3 d F = 412 df = 2 141 (B) 15 d F = 146df = 3 92 3 d F = 754 df = 3 116 Mean plusmn SEMleaves per plant (L) and plant height (H) in centimetersfor all treatments pooled were (A) L = 111 (04) H =176 (03) (B) L = 178 (04) H = 252 (04)

Table 1 Melon aphid control provided 15 and 3 dafter a single releasc of 34-42 convergent lady beetlcHper chrysanthemum pot

stored unmarked) (Fig 6B) However at 15 d af-ter release during July 1994 flown beetles provid-ed more control in comparison with stored beetles(Fig 6A) because flown beetles dispersed lessquickly (Fig 3A) so that more individuals werepresent to feed

Discussion

We found that preconditioning aggregation-col-lected H convergens or releasing reared beetles insome instances delayed dispersal in comparisonwith cold-stored beetles from aggregations Ourfindings differ from those of Starks et al (1975)who found no differences in dispersal after release

Mean SE controlMonth Year Treatmenti

15 d 3d

May 1993 All NS 345 (201)July 1993 All NS 784 43)Sept 1993 All NS 844 (59)July 1994 Flown 544 (111) 520 (64)Sept 1994 Flown 289 (147) 686 (63)July 1994 Stored 67 (160) 249 (129)Sept 1994 Stored 353 (116) 636 (58)Sept 1994 Unmarked 539 ( 70) 697 (69)

NS not sampled Calculated according to the Abbott (1925) methodb All all treatments (Hawn stored and reared) were released

on the same plants during 1993 so aphid consumption cannot beseparated by treatment and results from all treatments are com-bined

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 695

40

bull

Fi~ 7 Daily aphid consumption per adult convergentlady wtle (AB aphids per beetle mean t SEM) versusinitial density of melon aphids per leaf (AL aphids perleaf mean t SEM) on potted chrysantl1emums Theequation for the line is AB = 427 (SEM = 123) + 025(SEM = 005) AL (r2 = 082 F = 272 df = 1 6 P =0(02)

density-dependent functional response in preda-tion by H convergens has been demonstrated in amodel by Gutierrez et aI (1981) and has been re-ported for other aphid-feeding lady beetles in-cluding Cocci nella septempunctata L and Har-nwnia axyridis (Pallas) (Hukusima and Ohwaki1972 Hodek 1973)

Hippodamia convergens consumed a mean of 91(SEM = 94) aphids per day which agrees withHodeks (1973) conclusion that the daily feedingrate of adult aphidophagous Coccinellidae usuallyamounts to =100 aphids Clausen (1916) reared IIconvergens in the laboratory on hop aphids Pho-radon humuli (Schrank) which at maturity areabout the same size (1-2 mm body length) as mel-on aphids (Palmer 1952) Clausen found that adultH convergens over their life time consumed =66(SEM = 49 range 17-203) hop aphids each dayClausen used nearly full-grown aphids whereasours were of varying age and included smalleraphids which may account for the 13 lower num-ber of aphids eaten in Clausens (1916) study incomparison with our data Hagen and Sluss (1966)studied H convergens consumption of anothersmall species the spotted alfalfa aphid Therioa-phis maculata (Buckton) Preovipositional adultbeetles consumed 112 (SEM = 21) spotted alfalfaaphids per day and each beetle ate an average of91 aphids per day over their lifetime

Differences in aphid control or dispersal amongtreatments might be more dramatic with differentprerelease treatment of beetles Ignoffo et al(1977) found that clipping the wings of II conver-gens and other entomophages before release re-duced dispersal and increased predation Beetlesin our shldy were active during their 7-10 d out-doors in a screen tent before release (the flowntreatment) beetles moved with daily changes inshade and when disturbed during watering How-ever beetles spent little time actually flying in thetent The main difference between our treatmentswas that flown beetles were exposed to ambientlight wind temperature and humidity for 7-10 dbefore release in comparison with stored beetleswhich were kept at 4-1OdegC and constant dark untilthe day of release

With the exception of 1 lady beetle egg clusterin May 1993 (when reared beetles were released)no reproduction by H convergens was observedGiven rapid dispersal and lack of reproduction anycontrol from a single release would likely be tem-porary Releases are useful for inundative controlbut apparently are not effective for inoculative bi-ological control

Further research is needed to investigate the re-lationships among aphid density beetle releaserates and frequency dispersal and control Ourshldy plants harbored relatively high aphid densi-ties (Figs 5 and 6) and dispersal and control mightdiffer at the lower aphid densities more character-istic of a commercial nursery Hagen (1974) re-ports that the tendency of H convergens to feed

500100 200 300 400AphidsLeaf

oo

in sorghum of nondiapausing native H convergensin comparison with dispersal of aggregation-col-lected beetles Although preconditioning in ourstudy slowed dispersal treatments appear to havelittle practical benefit given that release of bothtreated and untreated beetles controlled aphids H(C)1lVlrglnsproduces few or no eggs unless fedaphids so rearing is expensive Preconditioning ina flight tent requires watering beetles and recol-lecting them before release Some beetles die dur-ing preconditioning reducing the number avail-able at the time of release Because H convergensis relatively inexpensive when purchased in bulk(=$40 buys 70000-80000 beetles) it is less ex-pensive to release greater numbers of stored bee-tles or to introduce them more often to compen-sate for their more rapid dispersal in comparisonwith preconditioning or rearing beetles

Although Starks et al (1975) found no differ-ences in dispersal among treatments native beetlesprovided better aphid control than aggregation-collected beetles When Starks et aI (1975) re-leased 2 beetles per plant in the greenhouse oncaged sorghum infested with 100-400 greenbugsSchizaphis graminum (Rondani) per plant theyobserved =95 and 67 aphid reduction respec-tively 5 d after release of native and aggregation-collected beetles These differences were appar-ently because beetles from aggregations oftenrested at the top of cages without feeding onaphids

Hippodamia convergens in our study consumed=25-170 melon aphids per day increasing con-sumption as aphid density increased (Fig 7) This

160 bull 19930 1994

Q) 120

fQ)Q)al-()C 80Ea4

696 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

aggregate and disperse varies with seasonal differ-ences in physiological condition of the beetlesHost plant aphid species release environmentand environmental conditions also may affectaphid control from beetle release

We are not aware of any studies indicating thatconvergent lady beetle releases are effective incontrolling aphids on a large scale Our preliminaryresearch on landscape roses indicates that releaseof relatively large numbers of beetles or multiplereleases may be needed to obtain control on largerplants Little is known about overwintering H con-vergens and the potential ecological effects oflarge-scale harvesting of aggregating lady beetleshave not been investigated Even if releases of ag-gregation-collected convergent lady beetles couldcontrol aphids on a large scale potential ecologicalor nontarget effects might make large-scale releas-es undesirable

There are quarantine considerations regardingthe shipment of field-collected insects Aggregat-ing beetles may harbor microorganisms and bulkcollections may be contaminated with other inver-tebrates The parasite Dinocampus (=Perilitus)coccinellae (Shrank) (Hymenoptera Braconidae)emerged from a few of the commercial H conver-gens we received We did not dissect hosts to checkparasitism and did not hold beetles and feed themaphids to induce D coccinellae emergence so wedo not know the extent of their parasitization byD coccinellae which attacks gt40 coccinellid spe-cies (Obrycki 1989) Ruzicka and Hagen (1985)found that 10 of the H convergens collectedfrom aggregations in California are parasitized byD coccinellae and parasitized beetles have less ofa tendency to disperse in comparison with unpar-asitized H convergens

Augmentative release of commercially availableconvergent lady beetles can provide inundativecontrol of relatively high aphid densities on sIllallpotted plants in a limited area Convergent ladybeetles are organically acceptable readily avail-able hardy easily handled and store well Pesti-cide exposure hazards and worker reentry restric-tions after pesticide application may make ladybeetle releases an attractive option for aphid con-trol in nurseries and greenhouses However ladybeetle releases must be integrated with other man-agement practices and pesticide applications forother pests may need to be modified for releasesto be effective

Acknowledgments

Michael P Parrella Norman E Gary Jill RentnerChuck LeGault and Bob Rousseau (Department of En-tomology University of California Davis) and Robin Ro-setta Pat Thompson and Ellen Zagory (Davis Arbore-tum University of California Davis) assisted usKenneth S Hagen and Kent M Daane (Laboratory ofBiological Control University of California Berkeley)Lester E Ehler (Department of Entomology Universityof California Davis) and Michael J Raupp (Department

of Entomology University of Maryland) critiqued draftsof the manuscript Unique Insect Control CitrusHeights CA provided the lady beetles The Universityof California Elvenia J Slosson Fund for OrnamentalHorticulture provided financial assistance

References Cited

Abbott W S 1925 A method of computing the ef-fectiveness of an insecticide J Econ Entomol 18265-267

Clausen C P 1916 Life-history and feeding recordsof a series of California Coccinellidae Univ CalifPub Entomol 1(6) 251-299

Cooke W C 1963 Ecology of the pea aphid in theblue mountain area of eastern Oregon and Washing-ton US Dep Agric Tech Bull 1287

Davidson W M 1919 The convergent ladybirdbeetle (Hippodamia convergens Guerin) and the bar-ley-com aphis (Aphis lIuzidis Fitch) Mon Bull CalifState Comm Homc 8 23-26

1924 Observations and experiments on the dispersionof the convergent lady-beetle (Hippodamia conver-gens Guerin-MenevilIe) in California Trans Am En-tomol Soc 50 163--175

Davis J R and R L Kirkland 1982 Physiologicaland environmental factors related to the dispersalflight of the convergent lady beetle Hippodamia con-vergens (Guerin-MenevilIe) J Kans Entomo Soc55 187-196

Eddy C O 1939 An attempt to colonize Hippodamiaconvergens Proc Int Soc Sugar Cane Technol 6385--386 (abstracted in Rev App Entomo 28 245[1940])

Fenton F A and R G Dahms 1951 Attempts atcontrolling the greenbug by the importation and re-lease of lady beetles in Oklahoma Proc Okla AcadSci 32 49-51

Flint M L S H Dreistadt J Rentner and M PParrella 1995 Lady beetle release controls aphidson potted plants Calif Agric 49(2) 5--8

Gutierrez A P J U Baumgaertner and K S Ha-gen 1981 A conceptual model for growth devel-opment and reproduction in the ladybird beetle Hip-podamia convergens (Coleoptera Coccinellidae)Can Entomo 113 21-33

Hagen K S 1962 Biology and ecology of predaceousCoccinellidae Annu Rev Entomol 7 289-326

1974 The significance of predaceous Coccinellidae inbiological and integrated control of insects Entomo-phaga 7 25-44

1982 Notes on the convergent lady beetle Univ CalifDiv Agric Nat Resour Leaf 2502

Hagen K S and R R Sluss 1966 Quantity ofaphids required for reproduction by Hippodamia sppin the laboratory pp 47-59 In I Hodek [ed] Ecol-ogy of aphidophagous insects Junk The Hague

Hatch M H and C Tanasse 1948 The liberationof Hippodamia convergens in the Yakima Valley ofWashington 1943 to 1946 J Econ EntomoI 41993

Hodek I 1973 Biology of Coccinellidae Junk TheHague

Hukusima S and T Ohwaki 1972 Further noteson feeding biology of Hannonia axyridis Pallas (Co-leoptera Coccinellidae) Res Bull Fac Agric GifuUniv 33 75--82

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 689

Campbell (1926) observed a decrease in peaaphids but did not credit this to the release be-cause it was too soon after liberation for progenyof the introduced beetles to have become effec-tive Recent research focused on dispersal and ef-fects of temperature wind and nutrition on dis-persal of H convergens (Hagen 1962 Rankin andRankin 1980 Davis and Kirkland 1982)

Despite long-standing opinions that H conver-gens releases are ineffective apparently onlyRaupp et aI (1994) and Flint et al (1995) haveconducted replicated controlled experiments onaphid density changes from release of H conver-gens Raupp et al (1994) released 10 H conver-gens per plant on potted firethorn Pyracantha le-landii infested with spirea aphid Aphis spiraecolaPatch and a woolly aphid (Eriosoma sp) in thegreenhouse Release controlled only spirea aphidsand only if phmts were caged An uncaged fieldrelease of1500 beetles by Raupp et al (1994) ap-parently reduced Capitophorus elaeagni (delCuerico) aphids on an Elaeagnus pungens Thun-berg hedge

Flint et aI (1995) reported significant reductionsin numbers of melon aphid Aphis gossypii Cloveron chrysanthemum Dendranthema grandifloraTzvelev and rose aphid Macrosiphum rosae (L)on rose Rosa sp when beetles were released out-doors on uncaged potted plants Here we reporton our investigation of whether dispersal of aggre-gation-collected H convergens was reduced by fly-ing and feeding beetles before releasing them Wecompared dispersal of insectary-reared beetleswith dispersal of beetles from aggregations and de-termined whetller inundative release of convergentlady beetles controls aphids on potted plants out-doors

Materials and Methods

We released marked adult convergent lady bee-tles outdoors on potted chrysanthemums infestedwith melon aphids We conducted 5 trials eachtrial of an experiment was conducted once duringMay July and September 1993 and July and Sep-tember 1994 Except for the insectary-reared bee-tles discussed below beetles were from Unique In-sect Control Citms Heights CA This commercialsupplier reports that beetles were from aggrega-tions at an altitude of 600 m in Placer CountyCalifornia and that beetles were collected 52 wkbefore each trial in July and September Beetlesreleased in May reportedly were collected in lateFebruary 1993 and stored at 4-100C

Treatment Effects on Beetle Dispersal Wepreconditioned (treated) beetles before releaseand conducted mark-release-recapture studies tocompare dispersal tendencies among treatmentsAll 5 trials included the following 2 treatmentsconducted with aggregation-collected lady beetles(1) beetles were allowed to fly for 1 wk (during1993) or 10 d (1994) in a screen-wall tent (3 by 3

by 2 m) outdoors sprayed daily with diluted honeyand marked before release (flown) or (2) beetleswere removed from cold storage (4-100C) 8 hbefore release confined witll honey and watermarked before release and sprayed with soda popimmediately after release (stored) Soda pop wasapplied because beetle distributors commonly rec-ommend application of this or other sweet liquidsto reduce beetle dispersal Third treatments wereadded to the trials during May 1993 and Septem-ber 1994 as described below

The May 1993 trial included a 3rd treatment ofbeetles produced in the insectary (reared) to de-termine whether rearing beetles influenced dis-persal in comparison with aggregation-collectedbeetles Reared H convergens adults were the Ist-generation offspring from adults collected in alfalfain Davis (Yolo County) in April Field-collectedadults and their progeny were held in petri dishesat 20degC and continuous light They were providedhoney water and ad lib an assortment of field-collected aphids mostly Aphis fabae Scopoli Aphisgossypii and green peach aphid Myzus persicae(Sulzer) After eclosion adults were held and fedthe same diet as larvae for 1-2 wk before release

The 3rd treatment during September 1994 wasdesigned to assess whether marking influencedbeetle dispersal Aggregation-collected beetleswere released but not marked before release (un-marked) They were otherwise handled the sameas stored beetles

To distinguish flown from stored beetles wemarked beetles before release by spraying 1 treat-ment with white Ace Instant Drying InteriorEx-terior Lacquer (Ace Hardware Oak Brook IL)Dark (black or blue) Ace Instant Drying InteriorExterior Lacquer was applied to beetles in the 2ndtreatment We also used aluminum Zynolyte SprayLacquer (Zynolyte Carson CA) during May whena 3rd middottreatment marking was needed

To apply lacquer we confined beetles from eachtreatment in a separate cardboard box (30 by 22by 5 cm) with window screen on 1 side andsprayed lacquer into the box through the screenonto beetles for 2 s We aired the box for 30 sto dry the lacquer and disperse solvent shook bee-tles down onto the bottom of the box allowed bee-tles to right themselves so that their elytra wereexposed and resprayed them We repeated thisspray-and-shake procedure 4 times until mostbeetles appeared to be marked

To assure ourselves of the effectiveness of mark-ing we retained and inspected beetles from treat-ments before release during July 1994 and Sep-tember 1993 and 1994 We collected 5-8 samplesof 17 (during 1993) or 40 (1994) beetles each fromevery treatment during each trial and examinedbeetles for marking using a dissecting binocularmicroscope Except for the unmarked treatmentduring September 1994 964 (SD = 46) of thebeetles recovered were marked We do not reportunmarked beetles recovered in the numbers pre-

690 ENVIRONMENTALENTOMOLOGY Vol 25 no 3

XX XX XX XX XXX 000 X 000 X 000 X 000 X 000X 000 X 000 X 000 X 000 X 000

XX XX XX XX XX

1993 10 m

xx xx xx xx xxX 000 X 000 X 000 X 000 X 000X 000 X 000 X 000 X 000 X 000

xx xx xx xx xx

XXXxoooo 0xoooo 0xxx

X X XxoooooxoooooX X X

X X Xxoooooxooooo

X X X

X X X X X X X X Xxooooo xooooo xoooooxooooo xooooo xooooo

X X X X X X X X XPlant Pots 1994 3m

X Rose x x x xxx xxxo Chrysanthemum Xooooo Xooooo XoooooXooooo xooooo xooooo

X X X X X X X X XFig 1 Convergent lady beetle experimental design Release and control plots were a randomized block 2 X 5

plots during 1993 (above within rectangle) and 3 x 3 plots during 1994 (below) except as described in Materialsand Methods Each 4-liter pot (0) contained 4 chrysanthemum plants with 6 (during 1993) or 10 (1994) pots perplot Each plot was sheltered on 3 sides by 6 (1993)or 8 (1994) aphid-free 1-m-tall rose Rosa sp Tropicana plantsin 20-liter pots (x) Beetle dispersal out of plots was evaluated by collecting all beetles from 1 or 2 previouslyunsampled pots in each release plot at intervals of 12 24 36 48 and 69 h after release Beetle movement withinplots was evaluated at each interval by resampling and removing beetles from all previously sampled pots

sented here except during September 1994 whenall unmarked beetles we recovered were presumedto be from the unmarked treatment beetles we re-leased

Beetle Releases We conducted all experimentsoutdoors at the University of California Davis Ar-boretum nursery in a gravel-surfaced area (130 by80 m) The area was free of plants except for thosein our study Each 4-liter pot contained 4 chrysan-themum plants We used 6 (1993) or 10 (1994)chrysanthemum pots (24 or 40 plants) in each plotsurrounded on 3 sides by potted roses Roses wereaphid-free and were used to shade the chrysanthe-mums which had been produced in a whitewashedgreenhouse Chrysanthemum pots did not touchwithin plots there was a gap (3-8 cm) between thecanopy of each pot Distance from plot edge to the

nearest other plots was 10 m (during 1993) or 3 In(1994) (Fig 1)

We counted the number of beetles to be re-leased into each pot by aspirating them into a vialselecting only apparently healthy (active) beetlesthat appeared marked We released beetles hyknocking them from vials onto potting media at thebase of the plants We released the stored beetlessprayed them with 10 ml of soda pop per potthen released the other treatment beetles in thatsame pot (during 1993) before moving to releasebeetles in the next pot Foliage and media werewet from watering plants immediately before re-lease

We released 34-42 beetles per pot (see h = 0in Results for exact numbers) once at dusk on 18May 20 July and 28 September 1993 and on 26

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 691

July and 27 September 1994 During 1993 we re-leased equal numbers of flown and stored H con-vergens in each pot During May when we alsonleased reared beetles 13 of the beetles releasedin each pot were of each treatment During 1994each release plot received only beetles from 1treatment either all flown or all stored beetles dur-ing July or all flown stored or unmarked beetlesduring September Release plots were replicated 6times during 1993 and 3 times for each treatmentduring 1994 as illustrated in Fig 1 Nonrelease(control) plots were replicated 4 (1993) or 3 (1994)times except that the May 1993 trial included only1 plot of control pots A randomized block of 2 X5 plots was used during 1993 3 X 3 during July1994 and 3 X 4 during September 1994 the for-mer 2 designs are illustrated in Fig 1 Before re-lease we controlled ants around all plants by de-ploying Grants Kills Ants (003 arsenic) (GrantLaboratories San Leandro CAl ant stakes

Beetle Dispersal At 12 24 36 48 and 69 hafter release we evaluated dispersal by aspiratingall beetles from 1 (during 1993) or 2 (1994) pre-viously unsampled pots in each plot Two excep-tions were as follows (1) during September 1993we collected beetles from 2 previously unsampledpots per plot at 69 h and (2) during July 1993 wecollected beetles at 12 24 and 48 h and beetleswere collected from 4 previously unsampled potsper plot at 48 h

Nondispersing beetles (none) were those recov-ered on previously unsampled pots and marked thesame color as the beetles released in that plot Wemeasured dispersal within plots at these same in-tervals by rcsampling and removing beetles fromall previously sampled pots for example at 36 hafter release during 1994 we removed all beetlesfrom 2 previously unsampled pots in each plot andremoved beetles that had moved onto pots sam-pled 12 and 24 h earlier To measure dispersalamong plots we recorded whether the beetles werecovered were marked with the same color as thebeetles released into that plot

Aphid Control We grew chrysanthemums inthe greenhouse under a photoperiod of 168 (LD)h at 20-30degC We planted 4 cuttings in each 4-literpot grew them for 8-10 wk then moved plantsoutdoors to the study area the day of release Weinoculated plants vith A gossypii 4--5wk after pot-ting and excluded other insects by screening yel-low sticky traps or removal of infested leaves

At ~36 h before release (to) we sampled aphidson 3 or 4 pots per plot during 1993 and on 4 or 6pots per plot during 1994 We resampled 2 potsper plot at 36 h after release during 1994 and re-sampled 2-4 pots per plot 72h after release dur-ing 1993 and 1994 The release plants sampled foraphids at 36 h were those from which we had justremoved beetles for the dispersal assessment de-scribed previously Release pots resampled 72 hafter release were those from which beetles were1st removed at 48 and 69 h after release except

that during September 1994 aphids were not sam-pled on pots from which beetles were first re-moved at 48 h

We counted all aphids on 1 leaf per plant (4leaves per pot) Each sample leaf was from abouthalfWayup the main stem from the soil line be-cause middle-canopy leaves consistently provide anaverage measure of aphid density per plant (Vehrset al 1992) During July 1994 we counted aphidson each 12 (on either side of the midvein) of 1middle canopy leaf on 32 plants and used a pairedt-test to determine whether there was significantdifference (t = 019 P = 085) in aphid densitybetween halves of the same leaf For the remainingplants during July and all plants during September1994 we counted aphids on only 12 of 1 leaf perplant and multiplied this value by 2 to obtainwhole-leaf estimates

In May 1993 we randomly allocated pots to re-lease or nonrelease plots During July and Septem-ber of both years before release we sampledaphids on 2 times the number of pots neededWe averaged the aphid density on the 4 plants perpot rank-ordered pots according to aphid densitydiscarded pots with the highest and lowest aphiddensities and allocated pots so that the averageprerelease aphid density was not significantly dif-ferent among plots We randomized pot locationwithin plots and randomly assigned plots to con-trols or releases

We counted the number of leaves and measuredthe height of 1 plant per pot We multiplied thenumber of leaves per plant by 4 plants per pot toestimate leaves per pot (LPP) We estimated dailyaphid consumption per beetle (AB)

AB = [ARto - ARtn) - (ACto - ACtn)] X LPP(Bto + Btn)2

where AR and AC are aphids per leaf on releaseand control plants respectively and B is beetlesper pot at release time (to) or 36 or 72 h afterrelease (tn) We included (AClo - ACtn) in theformula to adjust estimates of aphid consumptionThis adjustment was necessary because even in theabsence of lady beetles aphids declined during 4of 5 trials probably because of hot dry conditionsin the field or because some other aphid predatorscolonized plants after they were moved outdoorsor both

To determine whether there was a density-de-pendent functional response in predation we usedleast squares linear regression analyses (PROCREG SAS Institute 1988) to examine the relation-ship between aphid density at the time of release(ARto) (the independent variable) and daily aphidconsumption per beetle (AB) During 1993 we de-termined AB for all treatments combined duringeach of 3 trials (May July and September) be-cause all treatments were released into the sameplots Treatments were released in different plotsduring 1994 so we were able to calculate AB sep-

692 ENVIHONMENTAL ENTOMOLOGY Vol 25 no 3

Fig 2 Mean + SEM number of adult lady beetlesrecoveredafter a single release on chrysanlliemumsin-festedwillimelonaphidsfor trialsconductedin May(A)July (B) and September (e) 1993 Beetles were insec-tary-reared (reared) collected from aggregationsandcold-storeduntil the day of release (stored)or collectedfrom aggregationsand conditioned(flown)

beetles we recovered were in the same plots wherethey were released Because we also resampledplants from which we had previously removed allbeetles we found that when beetles did dispersethey usually left our study site (Fig 4 out cate-gory) rather ilian moving to nearby plants Most ofthe beetles iliat we did recover were on plants inthe pots where they were released (Fig 4 dis-persal category none) Fifteen percent or less ofthe beetles dispersed to oilier plants in ilieir re-lease plot (Fig 4 within category) Five percentor less of beetles were recovered in nearby plots(Fig 4 other category) Flown beetles weremore likely to disperse locally in comparison withstored beetles about twice as many flown beetleswere recovered on other plants (within category)in the release plot or in nearby plots (oilier cat-egory) (Fig 4) in comparison witll stored beetles

Marking beetles wiili lacquer to distinguishtreatments had no apparent effect on aphid con-

B

69

c

A

Reared

c=J Flown

- Stored

12 24 36 48

Hours After Release

o

20

10

10

20

o

0 0aCD 20c-ooOJ

~ampl 10-w(f)

-tIX 0

Results

Beetle Dispersal Most adult H convergens re-leased on groups of aphid-infested potted plants(Fig 1) dispersed within 1-3 d (Figs 2 and 3) In2 of 5 trials (Figs 2C and 3A) allowing these ag-gregation-collected beetles to drink and fly (flown)in a screen tent for 7-10 d before release delayeddispersal from plots in comparison with beetlesthat were stored at 4-lOoC (stored) until the dayof release About 20 (SE = 7) and 36 (SE =13) more flown beetles were recovered in com-parison with stored beetles during September 1993(Fig 2C) (t = 228 n = 36 P = 0029) and July1994 (Fig 3A) (t = 260 n = 90 P = 0011)respectively when samples from all postreleasetimes (12-69 h) were combined During July 1993significantly more stored than flown beetles wererecovered during all sample times pooled (Fig 2B)(t = 303 n = 36 P lt 0005)

Insectary-reared beetles dispersed from plotsmore slowly than did aggregation-collected beetles(Fig 2A)About 19 (SE = 10) and 41 (SE =9) more reared beetles were recovered in compar-ison wiili flown (t = 209 n = 30 P = 0046) orstored (t = 389 n = 30 P = 00005) beetlesrespectively when all postrelease samples werepooled during May 1993 (Fig 2A)

Because treatments were marked wiili a differ-ent color and released into separate plots during1994 we were able to determine that most of ilie

arately for each treatment during both 1994 trials(July and September) The result is 8 observationsfor the regression at 72 h There were 5 observa-tions for the regression at 36 h because aphidswere not sampled at that interval during 1993

Percentage control (C) was calculated for eachtreatment according to the Abbott (1925) method

C = [(N - R)N] X 100

where N is the percentage of aphids alive (postre-lease density divided by prerelease density multi-plied by 100) on nonrelease plants and R is thepercentage of aphids alive on release plants at 36or 72 h after release

Statistical Analyses For each treatment wecompared mean aphid densities before and afterreleases and mean numbers of beetles recoveredat all postrelease times combined Beetles in non-release plots were excluded from the data andanalyses Less than 1 of beetles recovered duringeach trial were collected in nonrelease plots Un-less stated otherwise we used t-tests for 2-waycomparisons and analysis of variance (PROCANOVA SAS Institute 1988) for multiple compar-isons Because the beetle dispersal out categoryequals the sum of all other categories subtractedfrom tlle total number of beetles released allmeans were not independent and multiple t-testsinstead of ANOVA were used to compare beetledispersal

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 693

Fig 3 Mean + SEM number of adult convergentlady beetles recovered after a single release on chrysan-themums infested with melon aphids for trials conductedin July (A) and September (B) 1994 Beetles were col-lected from aggregations and were cold-stored until theday of release (stored) cold-stored until release and notlacquered (unmarked) or conditioned (flown)

sumption (see below) or dispersal During Septem-ber 1994 we released cold-stored H convergensthat were both marked (stored) and not marked(unmarked) and there were no obvious differencesin average recovery at individual sample times(Fig 3B) or when all samples were pooled (t =068 n = 120 P = 050) No natural populationsof convergent lady beetles were observed on sur-rounding vegetation in September and presum-ably all unmarked beetles we recovered were thosewe released

Aphid Control Despite relatively rapid dispers-al from plots release of convergent lady beetlessignificantly reduced melon aphid densities on pot-ted chrysanthemums during all trials (Figs 5 and6) Within 3 d after a single release of 34-42 bee-tles per pot H convergens provided 25-84 con-trol as calculated according to the Abbott (1925)method Up to 54 control occurred 15 dafterrelease (Table 1)

Each adult H convergens consumed =25-170melon aphids per day (Fig 7) as estimated by di-viding the average reduction in aphids by the av-erage number of beetles present Samples collect-

A

c=J FloWn

- Storedc=J Unmarked

ILB

III eOther

B

A

c=J Flown

- Stored

Within

e e e f

d

Out

100

80-c~ 60Q)

gt8~ 40CIlQ) 20~Q)

CD-c 0Q)Vllaquol(I) 80Qj00 a~ 600

LiJCI) 40plusmn-IX

20

0None

ed 3 d after release reveal a strong positiveassociation between aphid density per leaf at thetime of release and daily aphid consumption perbeetle (AB) AB was higher on plants with a bigherinitial aphid density (Fig 7) Feeding rate and ini-tial aphid density were not signiRcantly correlatedat 15 d (r2 = 065 F = 56 df = 1 3 P = 01)aphids were sampled at this interval only during1994 and sample size was small

The extent of predation by poundlown beetles in com-parison with stored beetles could not be deter-mined during 1993 because equal numbers of bee-tles from all treatments were released on each potHowever poundlown and stored beetles were releasedinto separate plots during 1994 and both treat-ments significantly reduced aphid densities in com-parison with plots without beetles (Fig 6) Therewere no apparent differences in daily aphid con-sumption per beetle among treatments (poundlown

Dispersal

Fig 4 Mean + SEM dispersal (percentage of re-leased beetles recovered) for flownand stored convergentlady beetles for all sample times pooled over 3 d duringJuly (A) or September (B) 1994 In total 2400 beetleswere released in Julyand 2400 beetles (1200 each flownand stored) were released during September Nondis-persing beetles (none) were recovered in pots where theywere released Beetles that left the study area (out) werenot recovered Some beetles moved from the pot wherethey were released but remained within their release plot(within)The remaining beetles were recovered after dis-persing from release plants to other plots (other) in ourstudy area Within each month means that do not sharea common letter are significantlydifferent (P lt 005 t-tests)

6912 24 36 48Hours After Release

o

40

30

0 20abullQ) 10aCIlQ)Q)Q) 0enLiJ 40CI)Ix 30

20

10

0

694 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

c=J No Beetles - With Beetles A

0250 c==J No Beetles Stored B

- Flown ~ Unmarked200

150

100

50

0Before 15 Days 3 Days

500

-lt1l3 100

300

400

200

bulllt1lC()s1sclaquoWC)+-Ix

B

c

A

80

250200150100

50

- 0~ 50JQ 40a~ 30s~ 20gr 10+I 0

120

Fig 5 Mean + SEM melon aphids per leaf beforeand 3 d after release outdoors of 34-42 adult lady beetlesper pot each pot containing 4 chrysanthemum plants fortrials conducted in May (A) July (B) and September (C)1993 Aphid means in each trial were not significantlydifferent (P gt 001) before release (A) t = 139 n = 1272 (B) t = 031 n = 48 72 (C) t = 008 n = 64 96Means were significantly different (P lt 00001) after re-lease (A) t = 468 n = 20 72 (B) t = 1186 n = 4872 (C) t = 574 n = 6496 Mean plusmn SEM leaves perplant (L) and plant height (H) in centimeters for all treat-ments pooled were (A) L = 154 (04) H = 287 (O(l)(B) L = 158 (04) H = 301 (08) (C) L = 171 (06)H = 297 (11)

40

oBefore Release After Release

Fig 6 Mean + SEM melon aphids per leaf beforeand 15 and 3 d after release outdoors of 40 lown storedor unmarked adult convergent lady beetles per pot eachpot containing 4 chrysanthemum plants for trials con-ducted in July (A) and September (B) 1994 Aphid meansin each trial were not significantly different (P gt 005)before release (A) F = 00 df = 2 213 (B) F = 002df = 3 187 There were significant differences (P lt00001) in aphid densities after release of convergent ladybeetles ANOVA statistics are (A) 15 d F = 195 df =2 69 3 d F = 412 df = 2 141 (B) 15 d F = 146df = 3 92 3 d F = 754 df = 3 116 Mean plusmn SEMleaves per plant (L) and plant height (H) in centimetersfor all treatments pooled were (A) L = 111 (04) H =176 (03) (B) L = 178 (04) H = 252 (04)

Table 1 Melon aphid control provided 15 and 3 dafter a single releasc of 34-42 convergent lady beetlcHper chrysanthemum pot

stored unmarked) (Fig 6B) However at 15 d af-ter release during July 1994 flown beetles provid-ed more control in comparison with stored beetles(Fig 6A) because flown beetles dispersed lessquickly (Fig 3A) so that more individuals werepresent to feed

Discussion

We found that preconditioning aggregation-col-lected H convergens or releasing reared beetles insome instances delayed dispersal in comparisonwith cold-stored beetles from aggregations Ourfindings differ from those of Starks et al (1975)who found no differences in dispersal after release

Mean SE controlMonth Year Treatmenti

15 d 3d

May 1993 All NS 345 (201)July 1993 All NS 784 43)Sept 1993 All NS 844 (59)July 1994 Flown 544 (111) 520 (64)Sept 1994 Flown 289 (147) 686 (63)July 1994 Stored 67 (160) 249 (129)Sept 1994 Stored 353 (116) 636 (58)Sept 1994 Unmarked 539 ( 70) 697 (69)

NS not sampled Calculated according to the Abbott (1925) methodb All all treatments (Hawn stored and reared) were released

on the same plants during 1993 so aphid consumption cannot beseparated by treatment and results from all treatments are com-bined

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 695

40

bull

Fi~ 7 Daily aphid consumption per adult convergentlady wtle (AB aphids per beetle mean t SEM) versusinitial density of melon aphids per leaf (AL aphids perleaf mean t SEM) on potted chrysantl1emums Theequation for the line is AB = 427 (SEM = 123) + 025(SEM = 005) AL (r2 = 082 F = 272 df = 1 6 P =0(02)

density-dependent functional response in preda-tion by H convergens has been demonstrated in amodel by Gutierrez et aI (1981) and has been re-ported for other aphid-feeding lady beetles in-cluding Cocci nella septempunctata L and Har-nwnia axyridis (Pallas) (Hukusima and Ohwaki1972 Hodek 1973)

Hippodamia convergens consumed a mean of 91(SEM = 94) aphids per day which agrees withHodeks (1973) conclusion that the daily feedingrate of adult aphidophagous Coccinellidae usuallyamounts to =100 aphids Clausen (1916) reared IIconvergens in the laboratory on hop aphids Pho-radon humuli (Schrank) which at maturity areabout the same size (1-2 mm body length) as mel-on aphids (Palmer 1952) Clausen found that adultH convergens over their life time consumed =66(SEM = 49 range 17-203) hop aphids each dayClausen used nearly full-grown aphids whereasours were of varying age and included smalleraphids which may account for the 13 lower num-ber of aphids eaten in Clausens (1916) study incomparison with our data Hagen and Sluss (1966)studied H convergens consumption of anothersmall species the spotted alfalfa aphid Therioa-phis maculata (Buckton) Preovipositional adultbeetles consumed 112 (SEM = 21) spotted alfalfaaphids per day and each beetle ate an average of91 aphids per day over their lifetime

Differences in aphid control or dispersal amongtreatments might be more dramatic with differentprerelease treatment of beetles Ignoffo et al(1977) found that clipping the wings of II conver-gens and other entomophages before release re-duced dispersal and increased predation Beetlesin our shldy were active during their 7-10 d out-doors in a screen tent before release (the flowntreatment) beetles moved with daily changes inshade and when disturbed during watering How-ever beetles spent little time actually flying in thetent The main difference between our treatmentswas that flown beetles were exposed to ambientlight wind temperature and humidity for 7-10 dbefore release in comparison with stored beetleswhich were kept at 4-1OdegC and constant dark untilthe day of release

With the exception of 1 lady beetle egg clusterin May 1993 (when reared beetles were released)no reproduction by H convergens was observedGiven rapid dispersal and lack of reproduction anycontrol from a single release would likely be tem-porary Releases are useful for inundative controlbut apparently are not effective for inoculative bi-ological control

Further research is needed to investigate the re-lationships among aphid density beetle releaserates and frequency dispersal and control Ourshldy plants harbored relatively high aphid densi-ties (Figs 5 and 6) and dispersal and control mightdiffer at the lower aphid densities more character-istic of a commercial nursery Hagen (1974) re-ports that the tendency of H convergens to feed

500100 200 300 400AphidsLeaf

oo

in sorghum of nondiapausing native H convergensin comparison with dispersal of aggregation-col-lected beetles Although preconditioning in ourstudy slowed dispersal treatments appear to havelittle practical benefit given that release of bothtreated and untreated beetles controlled aphids H(C)1lVlrglnsproduces few or no eggs unless fedaphids so rearing is expensive Preconditioning ina flight tent requires watering beetles and recol-lecting them before release Some beetles die dur-ing preconditioning reducing the number avail-able at the time of release Because H convergensis relatively inexpensive when purchased in bulk(=$40 buys 70000-80000 beetles) it is less ex-pensive to release greater numbers of stored bee-tles or to introduce them more often to compen-sate for their more rapid dispersal in comparisonwith preconditioning or rearing beetles

Although Starks et al (1975) found no differ-ences in dispersal among treatments native beetlesprovided better aphid control than aggregation-collected beetles When Starks et aI (1975) re-leased 2 beetles per plant in the greenhouse oncaged sorghum infested with 100-400 greenbugsSchizaphis graminum (Rondani) per plant theyobserved =95 and 67 aphid reduction respec-tively 5 d after release of native and aggregation-collected beetles These differences were appar-ently because beetles from aggregations oftenrested at the top of cages without feeding onaphids

Hippodamia convergens in our study consumed=25-170 melon aphids per day increasing con-sumption as aphid density increased (Fig 7) This

160 bull 19930 1994

Q) 120

fQ)Q)al-()C 80Ea4

696 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

aggregate and disperse varies with seasonal differ-ences in physiological condition of the beetlesHost plant aphid species release environmentand environmental conditions also may affectaphid control from beetle release

We are not aware of any studies indicating thatconvergent lady beetle releases are effective incontrolling aphids on a large scale Our preliminaryresearch on landscape roses indicates that releaseof relatively large numbers of beetles or multiplereleases may be needed to obtain control on largerplants Little is known about overwintering H con-vergens and the potential ecological effects oflarge-scale harvesting of aggregating lady beetleshave not been investigated Even if releases of ag-gregation-collected convergent lady beetles couldcontrol aphids on a large scale potential ecologicalor nontarget effects might make large-scale releas-es undesirable

There are quarantine considerations regardingthe shipment of field-collected insects Aggregat-ing beetles may harbor microorganisms and bulkcollections may be contaminated with other inver-tebrates The parasite Dinocampus (=Perilitus)coccinellae (Shrank) (Hymenoptera Braconidae)emerged from a few of the commercial H conver-gens we received We did not dissect hosts to checkparasitism and did not hold beetles and feed themaphids to induce D coccinellae emergence so wedo not know the extent of their parasitization byD coccinellae which attacks gt40 coccinellid spe-cies (Obrycki 1989) Ruzicka and Hagen (1985)found that 10 of the H convergens collectedfrom aggregations in California are parasitized byD coccinellae and parasitized beetles have less ofa tendency to disperse in comparison with unpar-asitized H convergens

Augmentative release of commercially availableconvergent lady beetles can provide inundativecontrol of relatively high aphid densities on sIllallpotted plants in a limited area Convergent ladybeetles are organically acceptable readily avail-able hardy easily handled and store well Pesti-cide exposure hazards and worker reentry restric-tions after pesticide application may make ladybeetle releases an attractive option for aphid con-trol in nurseries and greenhouses However ladybeetle releases must be integrated with other man-agement practices and pesticide applications forother pests may need to be modified for releasesto be effective

Acknowledgments

Michael P Parrella Norman E Gary Jill RentnerChuck LeGault and Bob Rousseau (Department of En-tomology University of California Davis) and Robin Ro-setta Pat Thompson and Ellen Zagory (Davis Arbore-tum University of California Davis) assisted usKenneth S Hagen and Kent M Daane (Laboratory ofBiological Control University of California Berkeley)Lester E Ehler (Department of Entomology Universityof California Davis) and Michael J Raupp (Department

of Entomology University of Maryland) critiqued draftsof the manuscript Unique Insect Control CitrusHeights CA provided the lady beetles The Universityof California Elvenia J Slosson Fund for OrnamentalHorticulture provided financial assistance

References Cited

Abbott W S 1925 A method of computing the ef-fectiveness of an insecticide J Econ Entomol 18265-267

Clausen C P 1916 Life-history and feeding recordsof a series of California Coccinellidae Univ CalifPub Entomol 1(6) 251-299

Cooke W C 1963 Ecology of the pea aphid in theblue mountain area of eastern Oregon and Washing-ton US Dep Agric Tech Bull 1287

Davidson W M 1919 The convergent ladybirdbeetle (Hippodamia convergens Guerin) and the bar-ley-com aphis (Aphis lIuzidis Fitch) Mon Bull CalifState Comm Homc 8 23-26

1924 Observations and experiments on the dispersionof the convergent lady-beetle (Hippodamia conver-gens Guerin-MenevilIe) in California Trans Am En-tomol Soc 50 163--175

Davis J R and R L Kirkland 1982 Physiologicaland environmental factors related to the dispersalflight of the convergent lady beetle Hippodamia con-vergens (Guerin-MenevilIe) J Kans Entomo Soc55 187-196

Eddy C O 1939 An attempt to colonize Hippodamiaconvergens Proc Int Soc Sugar Cane Technol 6385--386 (abstracted in Rev App Entomo 28 245[1940])

Fenton F A and R G Dahms 1951 Attempts atcontrolling the greenbug by the importation and re-lease of lady beetles in Oklahoma Proc Okla AcadSci 32 49-51

Flint M L S H Dreistadt J Rentner and M PParrella 1995 Lady beetle release controls aphidson potted plants Calif Agric 49(2) 5--8

Gutierrez A P J U Baumgaertner and K S Ha-gen 1981 A conceptual model for growth devel-opment and reproduction in the ladybird beetle Hip-podamia convergens (Coleoptera Coccinellidae)Can Entomo 113 21-33

Hagen K S 1962 Biology and ecology of predaceousCoccinellidae Annu Rev Entomol 7 289-326

1974 The significance of predaceous Coccinellidae inbiological and integrated control of insects Entomo-phaga 7 25-44

1982 Notes on the convergent lady beetle Univ CalifDiv Agric Nat Resour Leaf 2502

Hagen K S and R R Sluss 1966 Quantity ofaphids required for reproduction by Hippodamia sppin the laboratory pp 47-59 In I Hodek [ed] Ecol-ogy of aphidophagous insects Junk The Hague

Hatch M H and C Tanasse 1948 The liberationof Hippodamia convergens in the Yakima Valley ofWashington 1943 to 1946 J Econ EntomoI 41993

Hodek I 1973 Biology of Coccinellidae Junk TheHague

Hukusima S and T Ohwaki 1972 Further noteson feeding biology of Hannonia axyridis Pallas (Co-leoptera Coccinellidae) Res Bull Fac Agric GifuUniv 33 75--82

690 ENVIRONMENTALENTOMOLOGY Vol 25 no 3

XX XX XX XX XXX 000 X 000 X 000 X 000 X 000X 000 X 000 X 000 X 000 X 000

XX XX XX XX XX

1993 10 m

xx xx xx xx xxX 000 X 000 X 000 X 000 X 000X 000 X 000 X 000 X 000 X 000

xx xx xx xx xx

XXXxoooo 0xoooo 0xxx

X X XxoooooxoooooX X X

X X Xxoooooxooooo

X X X

X X X X X X X X Xxooooo xooooo xoooooxooooo xooooo xooooo

X X X X X X X X XPlant Pots 1994 3m

X Rose x x x xxx xxxo Chrysanthemum Xooooo Xooooo XoooooXooooo xooooo xooooo

X X X X X X X X XFig 1 Convergent lady beetle experimental design Release and control plots were a randomized block 2 X 5

plots during 1993 (above within rectangle) and 3 x 3 plots during 1994 (below) except as described in Materialsand Methods Each 4-liter pot (0) contained 4 chrysanthemum plants with 6 (during 1993) or 10 (1994) pots perplot Each plot was sheltered on 3 sides by 6 (1993)or 8 (1994) aphid-free 1-m-tall rose Rosa sp Tropicana plantsin 20-liter pots (x) Beetle dispersal out of plots was evaluated by collecting all beetles from 1 or 2 previouslyunsampled pots in each release plot at intervals of 12 24 36 48 and 69 h after release Beetle movement withinplots was evaluated at each interval by resampling and removing beetles from all previously sampled pots

sented here except during September 1994 whenall unmarked beetles we recovered were presumedto be from the unmarked treatment beetles we re-leased

Beetle Releases We conducted all experimentsoutdoors at the University of California Davis Ar-boretum nursery in a gravel-surfaced area (130 by80 m) The area was free of plants except for thosein our study Each 4-liter pot contained 4 chrysan-themum plants We used 6 (1993) or 10 (1994)chrysanthemum pots (24 or 40 plants) in each plotsurrounded on 3 sides by potted roses Roses wereaphid-free and were used to shade the chrysanthe-mums which had been produced in a whitewashedgreenhouse Chrysanthemum pots did not touchwithin plots there was a gap (3-8 cm) between thecanopy of each pot Distance from plot edge to the

nearest other plots was 10 m (during 1993) or 3 In(1994) (Fig 1)

We counted the number of beetles to be re-leased into each pot by aspirating them into a vialselecting only apparently healthy (active) beetlesthat appeared marked We released beetles hyknocking them from vials onto potting media at thebase of the plants We released the stored beetlessprayed them with 10 ml of soda pop per potthen released the other treatment beetles in thatsame pot (during 1993) before moving to releasebeetles in the next pot Foliage and media werewet from watering plants immediately before re-lease

We released 34-42 beetles per pot (see h = 0in Results for exact numbers) once at dusk on 18May 20 July and 28 September 1993 and on 26

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 691

July and 27 September 1994 During 1993 we re-leased equal numbers of flown and stored H con-vergens in each pot During May when we alsonleased reared beetles 13 of the beetles releasedin each pot were of each treatment During 1994each release plot received only beetles from 1treatment either all flown or all stored beetles dur-ing July or all flown stored or unmarked beetlesduring September Release plots were replicated 6times during 1993 and 3 times for each treatmentduring 1994 as illustrated in Fig 1 Nonrelease(control) plots were replicated 4 (1993) or 3 (1994)times except that the May 1993 trial included only1 plot of control pots A randomized block of 2 X5 plots was used during 1993 3 X 3 during July1994 and 3 X 4 during September 1994 the for-mer 2 designs are illustrated in Fig 1 Before re-lease we controlled ants around all plants by de-ploying Grants Kills Ants (003 arsenic) (GrantLaboratories San Leandro CAl ant stakes

Beetle Dispersal At 12 24 36 48 and 69 hafter release we evaluated dispersal by aspiratingall beetles from 1 (during 1993) or 2 (1994) pre-viously unsampled pots in each plot Two excep-tions were as follows (1) during September 1993we collected beetles from 2 previously unsampledpots per plot at 69 h and (2) during July 1993 wecollected beetles at 12 24 and 48 h and beetleswere collected from 4 previously unsampled potsper plot at 48 h

Nondispersing beetles (none) were those recov-ered on previously unsampled pots and marked thesame color as the beetles released in that plot Wemeasured dispersal within plots at these same in-tervals by rcsampling and removing beetles fromall previously sampled pots for example at 36 hafter release during 1994 we removed all beetlesfrom 2 previously unsampled pots in each plot andremoved beetles that had moved onto pots sam-pled 12 and 24 h earlier To measure dispersalamong plots we recorded whether the beetles werecovered were marked with the same color as thebeetles released into that plot

Aphid Control We grew chrysanthemums inthe greenhouse under a photoperiod of 168 (LD)h at 20-30degC We planted 4 cuttings in each 4-literpot grew them for 8-10 wk then moved plantsoutdoors to the study area the day of release Weinoculated plants vith A gossypii 4--5wk after pot-ting and excluded other insects by screening yel-low sticky traps or removal of infested leaves

At ~36 h before release (to) we sampled aphidson 3 or 4 pots per plot during 1993 and on 4 or 6pots per plot during 1994 We resampled 2 potsper plot at 36 h after release during 1994 and re-sampled 2-4 pots per plot 72h after release dur-ing 1993 and 1994 The release plants sampled foraphids at 36 h were those from which we had justremoved beetles for the dispersal assessment de-scribed previously Release pots resampled 72 hafter release were those from which beetles were1st removed at 48 and 69 h after release except

that during September 1994 aphids were not sam-pled on pots from which beetles were first re-moved at 48 h

We counted all aphids on 1 leaf per plant (4leaves per pot) Each sample leaf was from abouthalfWayup the main stem from the soil line be-cause middle-canopy leaves consistently provide anaverage measure of aphid density per plant (Vehrset al 1992) During July 1994 we counted aphidson each 12 (on either side of the midvein) of 1middle canopy leaf on 32 plants and used a pairedt-test to determine whether there was significantdifference (t = 019 P = 085) in aphid densitybetween halves of the same leaf For the remainingplants during July and all plants during September1994 we counted aphids on only 12 of 1 leaf perplant and multiplied this value by 2 to obtainwhole-leaf estimates

In May 1993 we randomly allocated pots to re-lease or nonrelease plots During July and Septem-ber of both years before release we sampledaphids on 2 times the number of pots neededWe averaged the aphid density on the 4 plants perpot rank-ordered pots according to aphid densitydiscarded pots with the highest and lowest aphiddensities and allocated pots so that the averageprerelease aphid density was not significantly dif-ferent among plots We randomized pot locationwithin plots and randomly assigned plots to con-trols or releases

We counted the number of leaves and measuredthe height of 1 plant per pot We multiplied thenumber of leaves per plant by 4 plants per pot toestimate leaves per pot (LPP) We estimated dailyaphid consumption per beetle (AB)

AB = [ARto - ARtn) - (ACto - ACtn)] X LPP(Bto + Btn)2

where AR and AC are aphids per leaf on releaseand control plants respectively and B is beetlesper pot at release time (to) or 36 or 72 h afterrelease (tn) We included (AClo - ACtn) in theformula to adjust estimates of aphid consumptionThis adjustment was necessary because even in theabsence of lady beetles aphids declined during 4of 5 trials probably because of hot dry conditionsin the field or because some other aphid predatorscolonized plants after they were moved outdoorsor both

To determine whether there was a density-de-pendent functional response in predation we usedleast squares linear regression analyses (PROCREG SAS Institute 1988) to examine the relation-ship between aphid density at the time of release(ARto) (the independent variable) and daily aphidconsumption per beetle (AB) During 1993 we de-termined AB for all treatments combined duringeach of 3 trials (May July and September) be-cause all treatments were released into the sameplots Treatments were released in different plotsduring 1994 so we were able to calculate AB sep-

692 ENVIHONMENTAL ENTOMOLOGY Vol 25 no 3

Fig 2 Mean + SEM number of adult lady beetlesrecoveredafter a single release on chrysanlliemumsin-festedwillimelonaphidsfor trialsconductedin May(A)July (B) and September (e) 1993 Beetles were insec-tary-reared (reared) collected from aggregationsandcold-storeduntil the day of release (stored)or collectedfrom aggregationsand conditioned(flown)

beetles we recovered were in the same plots wherethey were released Because we also resampledplants from which we had previously removed allbeetles we found that when beetles did dispersethey usually left our study site (Fig 4 out cate-gory) rather ilian moving to nearby plants Most ofthe beetles iliat we did recover were on plants inthe pots where they were released (Fig 4 dis-persal category none) Fifteen percent or less ofthe beetles dispersed to oilier plants in ilieir re-lease plot (Fig 4 within category) Five percentor less of beetles were recovered in nearby plots(Fig 4 other category) Flown beetles weremore likely to disperse locally in comparison withstored beetles about twice as many flown beetleswere recovered on other plants (within category)in the release plot or in nearby plots (oilier cat-egory) (Fig 4) in comparison witll stored beetles

Marking beetles wiili lacquer to distinguishtreatments had no apparent effect on aphid con-

B

69

c

A

Reared

c=J Flown

- Stored

12 24 36 48

Hours After Release

o

20

10

10

20

o

0 0aCD 20c-ooOJ

~ampl 10-w(f)

-tIX 0

Results

Beetle Dispersal Most adult H convergens re-leased on groups of aphid-infested potted plants(Fig 1) dispersed within 1-3 d (Figs 2 and 3) In2 of 5 trials (Figs 2C and 3A) allowing these ag-gregation-collected beetles to drink and fly (flown)in a screen tent for 7-10 d before release delayeddispersal from plots in comparison with beetlesthat were stored at 4-lOoC (stored) until the dayof release About 20 (SE = 7) and 36 (SE =13) more flown beetles were recovered in com-parison with stored beetles during September 1993(Fig 2C) (t = 228 n = 36 P = 0029) and July1994 (Fig 3A) (t = 260 n = 90 P = 0011)respectively when samples from all postreleasetimes (12-69 h) were combined During July 1993significantly more stored than flown beetles wererecovered during all sample times pooled (Fig 2B)(t = 303 n = 36 P lt 0005)

Insectary-reared beetles dispersed from plotsmore slowly than did aggregation-collected beetles(Fig 2A)About 19 (SE = 10) and 41 (SE =9) more reared beetles were recovered in compar-ison wiili flown (t = 209 n = 30 P = 0046) orstored (t = 389 n = 30 P = 00005) beetlesrespectively when all postrelease samples werepooled during May 1993 (Fig 2A)

Because treatments were marked wiili a differ-ent color and released into separate plots during1994 we were able to determine that most of ilie

arately for each treatment during both 1994 trials(July and September) The result is 8 observationsfor the regression at 72 h There were 5 observa-tions for the regression at 36 h because aphidswere not sampled at that interval during 1993

Percentage control (C) was calculated for eachtreatment according to the Abbott (1925) method

C = [(N - R)N] X 100

where N is the percentage of aphids alive (postre-lease density divided by prerelease density multi-plied by 100) on nonrelease plants and R is thepercentage of aphids alive on release plants at 36or 72 h after release

Statistical Analyses For each treatment wecompared mean aphid densities before and afterreleases and mean numbers of beetles recoveredat all postrelease times combined Beetles in non-release plots were excluded from the data andanalyses Less than 1 of beetles recovered duringeach trial were collected in nonrelease plots Un-less stated otherwise we used t-tests for 2-waycomparisons and analysis of variance (PROCANOVA SAS Institute 1988) for multiple compar-isons Because the beetle dispersal out categoryequals the sum of all other categories subtractedfrom tlle total number of beetles released allmeans were not independent and multiple t-testsinstead of ANOVA were used to compare beetledispersal

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 693

Fig 3 Mean + SEM number of adult convergentlady beetles recovered after a single release on chrysan-themums infested with melon aphids for trials conductedin July (A) and September (B) 1994 Beetles were col-lected from aggregations and were cold-stored until theday of release (stored) cold-stored until release and notlacquered (unmarked) or conditioned (flown)

sumption (see below) or dispersal During Septem-ber 1994 we released cold-stored H convergensthat were both marked (stored) and not marked(unmarked) and there were no obvious differencesin average recovery at individual sample times(Fig 3B) or when all samples were pooled (t =068 n = 120 P = 050) No natural populationsof convergent lady beetles were observed on sur-rounding vegetation in September and presum-ably all unmarked beetles we recovered were thosewe released

Aphid Control Despite relatively rapid dispers-al from plots release of convergent lady beetlessignificantly reduced melon aphid densities on pot-ted chrysanthemums during all trials (Figs 5 and6) Within 3 d after a single release of 34-42 bee-tles per pot H convergens provided 25-84 con-trol as calculated according to the Abbott (1925)method Up to 54 control occurred 15 dafterrelease (Table 1)

Each adult H convergens consumed =25-170melon aphids per day (Fig 7) as estimated by di-viding the average reduction in aphids by the av-erage number of beetles present Samples collect-

A

c=J FloWn

- Storedc=J Unmarked

ILB

III eOther

B

A

c=J Flown

- Stored

Within

e e e f

d

Out

100

80-c~ 60Q)

gt8~ 40CIlQ) 20~Q)

CD-c 0Q)Vllaquol(I) 80Qj00 a~ 600

LiJCI) 40plusmn-IX

20

0None

ed 3 d after release reveal a strong positiveassociation between aphid density per leaf at thetime of release and daily aphid consumption perbeetle (AB) AB was higher on plants with a bigherinitial aphid density (Fig 7) Feeding rate and ini-tial aphid density were not signiRcantly correlatedat 15 d (r2 = 065 F = 56 df = 1 3 P = 01)aphids were sampled at this interval only during1994 and sample size was small

The extent of predation by poundlown beetles in com-parison with stored beetles could not be deter-mined during 1993 because equal numbers of bee-tles from all treatments were released on each potHowever poundlown and stored beetles were releasedinto separate plots during 1994 and both treat-ments significantly reduced aphid densities in com-parison with plots without beetles (Fig 6) Therewere no apparent differences in daily aphid con-sumption per beetle among treatments (poundlown

Dispersal

Fig 4 Mean + SEM dispersal (percentage of re-leased beetles recovered) for flownand stored convergentlady beetles for all sample times pooled over 3 d duringJuly (A) or September (B) 1994 In total 2400 beetleswere released in Julyand 2400 beetles (1200 each flownand stored) were released during September Nondis-persing beetles (none) were recovered in pots where theywere released Beetles that left the study area (out) werenot recovered Some beetles moved from the pot wherethey were released but remained within their release plot(within)The remaining beetles were recovered after dis-persing from release plants to other plots (other) in ourstudy area Within each month means that do not sharea common letter are significantlydifferent (P lt 005 t-tests)

6912 24 36 48Hours After Release

o

40

30

0 20abullQ) 10aCIlQ)Q)Q) 0enLiJ 40CI)Ix 30

20

10

0

694 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

c=J No Beetles - With Beetles A

0250 c==J No Beetles Stored B

- Flown ~ Unmarked200

150

100

50

0Before 15 Days 3 Days

500

-lt1l3 100

300

400

200

bulllt1lC()s1sclaquoWC)+-Ix

B

c

A

80

250200150100

50

- 0~ 50JQ 40a~ 30s~ 20gr 10+I 0

120

Fig 5 Mean + SEM melon aphids per leaf beforeand 3 d after release outdoors of 34-42 adult lady beetlesper pot each pot containing 4 chrysanthemum plants fortrials conducted in May (A) July (B) and September (C)1993 Aphid means in each trial were not significantlydifferent (P gt 001) before release (A) t = 139 n = 1272 (B) t = 031 n = 48 72 (C) t = 008 n = 64 96Means were significantly different (P lt 00001) after re-lease (A) t = 468 n = 20 72 (B) t = 1186 n = 4872 (C) t = 574 n = 6496 Mean plusmn SEM leaves perplant (L) and plant height (H) in centimeters for all treat-ments pooled were (A) L = 154 (04) H = 287 (O(l)(B) L = 158 (04) H = 301 (08) (C) L = 171 (06)H = 297 (11)

40

oBefore Release After Release

Fig 6 Mean + SEM melon aphids per leaf beforeand 15 and 3 d after release outdoors of 40 lown storedor unmarked adult convergent lady beetles per pot eachpot containing 4 chrysanthemum plants for trials con-ducted in July (A) and September (B) 1994 Aphid meansin each trial were not significantly different (P gt 005)before release (A) F = 00 df = 2 213 (B) F = 002df = 3 187 There were significant differences (P lt00001) in aphid densities after release of convergent ladybeetles ANOVA statistics are (A) 15 d F = 195 df =2 69 3 d F = 412 df = 2 141 (B) 15 d F = 146df = 3 92 3 d F = 754 df = 3 116 Mean plusmn SEMleaves per plant (L) and plant height (H) in centimetersfor all treatments pooled were (A) L = 111 (04) H =176 (03) (B) L = 178 (04) H = 252 (04)

Table 1 Melon aphid control provided 15 and 3 dafter a single releasc of 34-42 convergent lady beetlcHper chrysanthemum pot

stored unmarked) (Fig 6B) However at 15 d af-ter release during July 1994 flown beetles provid-ed more control in comparison with stored beetles(Fig 6A) because flown beetles dispersed lessquickly (Fig 3A) so that more individuals werepresent to feed

Discussion

We found that preconditioning aggregation-col-lected H convergens or releasing reared beetles insome instances delayed dispersal in comparisonwith cold-stored beetles from aggregations Ourfindings differ from those of Starks et al (1975)who found no differences in dispersal after release

Mean SE controlMonth Year Treatmenti

15 d 3d

May 1993 All NS 345 (201)July 1993 All NS 784 43)Sept 1993 All NS 844 (59)July 1994 Flown 544 (111) 520 (64)Sept 1994 Flown 289 (147) 686 (63)July 1994 Stored 67 (160) 249 (129)Sept 1994 Stored 353 (116) 636 (58)Sept 1994 Unmarked 539 ( 70) 697 (69)

NS not sampled Calculated according to the Abbott (1925) methodb All all treatments (Hawn stored and reared) were released

on the same plants during 1993 so aphid consumption cannot beseparated by treatment and results from all treatments are com-bined

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 695

40

bull

Fi~ 7 Daily aphid consumption per adult convergentlady wtle (AB aphids per beetle mean t SEM) versusinitial density of melon aphids per leaf (AL aphids perleaf mean t SEM) on potted chrysantl1emums Theequation for the line is AB = 427 (SEM = 123) + 025(SEM = 005) AL (r2 = 082 F = 272 df = 1 6 P =0(02)

density-dependent functional response in preda-tion by H convergens has been demonstrated in amodel by Gutierrez et aI (1981) and has been re-ported for other aphid-feeding lady beetles in-cluding Cocci nella septempunctata L and Har-nwnia axyridis (Pallas) (Hukusima and Ohwaki1972 Hodek 1973)

Hippodamia convergens consumed a mean of 91(SEM = 94) aphids per day which agrees withHodeks (1973) conclusion that the daily feedingrate of adult aphidophagous Coccinellidae usuallyamounts to =100 aphids Clausen (1916) reared IIconvergens in the laboratory on hop aphids Pho-radon humuli (Schrank) which at maturity areabout the same size (1-2 mm body length) as mel-on aphids (Palmer 1952) Clausen found that adultH convergens over their life time consumed =66(SEM = 49 range 17-203) hop aphids each dayClausen used nearly full-grown aphids whereasours were of varying age and included smalleraphids which may account for the 13 lower num-ber of aphids eaten in Clausens (1916) study incomparison with our data Hagen and Sluss (1966)studied H convergens consumption of anothersmall species the spotted alfalfa aphid Therioa-phis maculata (Buckton) Preovipositional adultbeetles consumed 112 (SEM = 21) spotted alfalfaaphids per day and each beetle ate an average of91 aphids per day over their lifetime

Differences in aphid control or dispersal amongtreatments might be more dramatic with differentprerelease treatment of beetles Ignoffo et al(1977) found that clipping the wings of II conver-gens and other entomophages before release re-duced dispersal and increased predation Beetlesin our shldy were active during their 7-10 d out-doors in a screen tent before release (the flowntreatment) beetles moved with daily changes inshade and when disturbed during watering How-ever beetles spent little time actually flying in thetent The main difference between our treatmentswas that flown beetles were exposed to ambientlight wind temperature and humidity for 7-10 dbefore release in comparison with stored beetleswhich were kept at 4-1OdegC and constant dark untilthe day of release

With the exception of 1 lady beetle egg clusterin May 1993 (when reared beetles were released)no reproduction by H convergens was observedGiven rapid dispersal and lack of reproduction anycontrol from a single release would likely be tem-porary Releases are useful for inundative controlbut apparently are not effective for inoculative bi-ological control

Further research is needed to investigate the re-lationships among aphid density beetle releaserates and frequency dispersal and control Ourshldy plants harbored relatively high aphid densi-ties (Figs 5 and 6) and dispersal and control mightdiffer at the lower aphid densities more character-istic of a commercial nursery Hagen (1974) re-ports that the tendency of H convergens to feed

500100 200 300 400AphidsLeaf

oo

in sorghum of nondiapausing native H convergensin comparison with dispersal of aggregation-col-lected beetles Although preconditioning in ourstudy slowed dispersal treatments appear to havelittle practical benefit given that release of bothtreated and untreated beetles controlled aphids H(C)1lVlrglnsproduces few or no eggs unless fedaphids so rearing is expensive Preconditioning ina flight tent requires watering beetles and recol-lecting them before release Some beetles die dur-ing preconditioning reducing the number avail-able at the time of release Because H convergensis relatively inexpensive when purchased in bulk(=$40 buys 70000-80000 beetles) it is less ex-pensive to release greater numbers of stored bee-tles or to introduce them more often to compen-sate for their more rapid dispersal in comparisonwith preconditioning or rearing beetles

Although Starks et al (1975) found no differ-ences in dispersal among treatments native beetlesprovided better aphid control than aggregation-collected beetles When Starks et aI (1975) re-leased 2 beetles per plant in the greenhouse oncaged sorghum infested with 100-400 greenbugsSchizaphis graminum (Rondani) per plant theyobserved =95 and 67 aphid reduction respec-tively 5 d after release of native and aggregation-collected beetles These differences were appar-ently because beetles from aggregations oftenrested at the top of cages without feeding onaphids

Hippodamia convergens in our study consumed=25-170 melon aphids per day increasing con-sumption as aphid density increased (Fig 7) This

160 bull 19930 1994

Q) 120

fQ)Q)al-()C 80Ea4

696 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

aggregate and disperse varies with seasonal differ-ences in physiological condition of the beetlesHost plant aphid species release environmentand environmental conditions also may affectaphid control from beetle release

We are not aware of any studies indicating thatconvergent lady beetle releases are effective incontrolling aphids on a large scale Our preliminaryresearch on landscape roses indicates that releaseof relatively large numbers of beetles or multiplereleases may be needed to obtain control on largerplants Little is known about overwintering H con-vergens and the potential ecological effects oflarge-scale harvesting of aggregating lady beetleshave not been investigated Even if releases of ag-gregation-collected convergent lady beetles couldcontrol aphids on a large scale potential ecologicalor nontarget effects might make large-scale releas-es undesirable

There are quarantine considerations regardingthe shipment of field-collected insects Aggregat-ing beetles may harbor microorganisms and bulkcollections may be contaminated with other inver-tebrates The parasite Dinocampus (=Perilitus)coccinellae (Shrank) (Hymenoptera Braconidae)emerged from a few of the commercial H conver-gens we received We did not dissect hosts to checkparasitism and did not hold beetles and feed themaphids to induce D coccinellae emergence so wedo not know the extent of their parasitization byD coccinellae which attacks gt40 coccinellid spe-cies (Obrycki 1989) Ruzicka and Hagen (1985)found that 10 of the H convergens collectedfrom aggregations in California are parasitized byD coccinellae and parasitized beetles have less ofa tendency to disperse in comparison with unpar-asitized H convergens

Augmentative release of commercially availableconvergent lady beetles can provide inundativecontrol of relatively high aphid densities on sIllallpotted plants in a limited area Convergent ladybeetles are organically acceptable readily avail-able hardy easily handled and store well Pesti-cide exposure hazards and worker reentry restric-tions after pesticide application may make ladybeetle releases an attractive option for aphid con-trol in nurseries and greenhouses However ladybeetle releases must be integrated with other man-agement practices and pesticide applications forother pests may need to be modified for releasesto be effective

Acknowledgments

Michael P Parrella Norman E Gary Jill RentnerChuck LeGault and Bob Rousseau (Department of En-tomology University of California Davis) and Robin Ro-setta Pat Thompson and Ellen Zagory (Davis Arbore-tum University of California Davis) assisted usKenneth S Hagen and Kent M Daane (Laboratory ofBiological Control University of California Berkeley)Lester E Ehler (Department of Entomology Universityof California Davis) and Michael J Raupp (Department

of Entomology University of Maryland) critiqued draftsof the manuscript Unique Insect Control CitrusHeights CA provided the lady beetles The Universityof California Elvenia J Slosson Fund for OrnamentalHorticulture provided financial assistance

References Cited

Abbott W S 1925 A method of computing the ef-fectiveness of an insecticide J Econ Entomol 18265-267

Clausen C P 1916 Life-history and feeding recordsof a series of California Coccinellidae Univ CalifPub Entomol 1(6) 251-299

Cooke W C 1963 Ecology of the pea aphid in theblue mountain area of eastern Oregon and Washing-ton US Dep Agric Tech Bull 1287

Davidson W M 1919 The convergent ladybirdbeetle (Hippodamia convergens Guerin) and the bar-ley-com aphis (Aphis lIuzidis Fitch) Mon Bull CalifState Comm Homc 8 23-26

1924 Observations and experiments on the dispersionof the convergent lady-beetle (Hippodamia conver-gens Guerin-MenevilIe) in California Trans Am En-tomol Soc 50 163--175

Davis J R and R L Kirkland 1982 Physiologicaland environmental factors related to the dispersalflight of the convergent lady beetle Hippodamia con-vergens (Guerin-MenevilIe) J Kans Entomo Soc55 187-196

Eddy C O 1939 An attempt to colonize Hippodamiaconvergens Proc Int Soc Sugar Cane Technol 6385--386 (abstracted in Rev App Entomo 28 245[1940])

Fenton F A and R G Dahms 1951 Attempts atcontrolling the greenbug by the importation and re-lease of lady beetles in Oklahoma Proc Okla AcadSci 32 49-51

Flint M L S H Dreistadt J Rentner and M PParrella 1995 Lady beetle release controls aphidson potted plants Calif Agric 49(2) 5--8

Gutierrez A P J U Baumgaertner and K S Ha-gen 1981 A conceptual model for growth devel-opment and reproduction in the ladybird beetle Hip-podamia convergens (Coleoptera Coccinellidae)Can Entomo 113 21-33

Hagen K S 1962 Biology and ecology of predaceousCoccinellidae Annu Rev Entomol 7 289-326

1974 The significance of predaceous Coccinellidae inbiological and integrated control of insects Entomo-phaga 7 25-44

1982 Notes on the convergent lady beetle Univ CalifDiv Agric Nat Resour Leaf 2502

Hagen K S and R R Sluss 1966 Quantity ofaphids required for reproduction by Hippodamia sppin the laboratory pp 47-59 In I Hodek [ed] Ecol-ogy of aphidophagous insects Junk The Hague

Hatch M H and C Tanasse 1948 The liberationof Hippodamia convergens in the Yakima Valley ofWashington 1943 to 1946 J Econ EntomoI 41993

Hodek I 1973 Biology of Coccinellidae Junk TheHague

Hukusima S and T Ohwaki 1972 Further noteson feeding biology of Hannonia axyridis Pallas (Co-leoptera Coccinellidae) Res Bull Fac Agric GifuUniv 33 75--82

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 691

July and 27 September 1994 During 1993 we re-leased equal numbers of flown and stored H con-vergens in each pot During May when we alsonleased reared beetles 13 of the beetles releasedin each pot were of each treatment During 1994each release plot received only beetles from 1treatment either all flown or all stored beetles dur-ing July or all flown stored or unmarked beetlesduring September Release plots were replicated 6times during 1993 and 3 times for each treatmentduring 1994 as illustrated in Fig 1 Nonrelease(control) plots were replicated 4 (1993) or 3 (1994)times except that the May 1993 trial included only1 plot of control pots A randomized block of 2 X5 plots was used during 1993 3 X 3 during July1994 and 3 X 4 during September 1994 the for-mer 2 designs are illustrated in Fig 1 Before re-lease we controlled ants around all plants by de-ploying Grants Kills Ants (003 arsenic) (GrantLaboratories San Leandro CAl ant stakes

Beetle Dispersal At 12 24 36 48 and 69 hafter release we evaluated dispersal by aspiratingall beetles from 1 (during 1993) or 2 (1994) pre-viously unsampled pots in each plot Two excep-tions were as follows (1) during September 1993we collected beetles from 2 previously unsampledpots per plot at 69 h and (2) during July 1993 wecollected beetles at 12 24 and 48 h and beetleswere collected from 4 previously unsampled potsper plot at 48 h

Nondispersing beetles (none) were those recov-ered on previously unsampled pots and marked thesame color as the beetles released in that plot Wemeasured dispersal within plots at these same in-tervals by rcsampling and removing beetles fromall previously sampled pots for example at 36 hafter release during 1994 we removed all beetlesfrom 2 previously unsampled pots in each plot andremoved beetles that had moved onto pots sam-pled 12 and 24 h earlier To measure dispersalamong plots we recorded whether the beetles werecovered were marked with the same color as thebeetles released into that plot

Aphid Control We grew chrysanthemums inthe greenhouse under a photoperiod of 168 (LD)h at 20-30degC We planted 4 cuttings in each 4-literpot grew them for 8-10 wk then moved plantsoutdoors to the study area the day of release Weinoculated plants vith A gossypii 4--5wk after pot-ting and excluded other insects by screening yel-low sticky traps or removal of infested leaves

At ~36 h before release (to) we sampled aphidson 3 or 4 pots per plot during 1993 and on 4 or 6pots per plot during 1994 We resampled 2 potsper plot at 36 h after release during 1994 and re-sampled 2-4 pots per plot 72h after release dur-ing 1993 and 1994 The release plants sampled foraphids at 36 h were those from which we had justremoved beetles for the dispersal assessment de-scribed previously Release pots resampled 72 hafter release were those from which beetles were1st removed at 48 and 69 h after release except

that during September 1994 aphids were not sam-pled on pots from which beetles were first re-moved at 48 h

We counted all aphids on 1 leaf per plant (4leaves per pot) Each sample leaf was from abouthalfWayup the main stem from the soil line be-cause middle-canopy leaves consistently provide anaverage measure of aphid density per plant (Vehrset al 1992) During July 1994 we counted aphidson each 12 (on either side of the midvein) of 1middle canopy leaf on 32 plants and used a pairedt-test to determine whether there was significantdifference (t = 019 P = 085) in aphid densitybetween halves of the same leaf For the remainingplants during July and all plants during September1994 we counted aphids on only 12 of 1 leaf perplant and multiplied this value by 2 to obtainwhole-leaf estimates

In May 1993 we randomly allocated pots to re-lease or nonrelease plots During July and Septem-ber of both years before release we sampledaphids on 2 times the number of pots neededWe averaged the aphid density on the 4 plants perpot rank-ordered pots according to aphid densitydiscarded pots with the highest and lowest aphiddensities and allocated pots so that the averageprerelease aphid density was not significantly dif-ferent among plots We randomized pot locationwithin plots and randomly assigned plots to con-trols or releases

We counted the number of leaves and measuredthe height of 1 plant per pot We multiplied thenumber of leaves per plant by 4 plants per pot toestimate leaves per pot (LPP) We estimated dailyaphid consumption per beetle (AB)

AB = [ARto - ARtn) - (ACto - ACtn)] X LPP(Bto + Btn)2

where AR and AC are aphids per leaf on releaseand control plants respectively and B is beetlesper pot at release time (to) or 36 or 72 h afterrelease (tn) We included (AClo - ACtn) in theformula to adjust estimates of aphid consumptionThis adjustment was necessary because even in theabsence of lady beetles aphids declined during 4of 5 trials probably because of hot dry conditionsin the field or because some other aphid predatorscolonized plants after they were moved outdoorsor both

To determine whether there was a density-de-pendent functional response in predation we usedleast squares linear regression analyses (PROCREG SAS Institute 1988) to examine the relation-ship between aphid density at the time of release(ARto) (the independent variable) and daily aphidconsumption per beetle (AB) During 1993 we de-termined AB for all treatments combined duringeach of 3 trials (May July and September) be-cause all treatments were released into the sameplots Treatments were released in different plotsduring 1994 so we were able to calculate AB sep-

692 ENVIHONMENTAL ENTOMOLOGY Vol 25 no 3

Fig 2 Mean + SEM number of adult lady beetlesrecoveredafter a single release on chrysanlliemumsin-festedwillimelonaphidsfor trialsconductedin May(A)July (B) and September (e) 1993 Beetles were insec-tary-reared (reared) collected from aggregationsandcold-storeduntil the day of release (stored)or collectedfrom aggregationsand conditioned(flown)

beetles we recovered were in the same plots wherethey were released Because we also resampledplants from which we had previously removed allbeetles we found that when beetles did dispersethey usually left our study site (Fig 4 out cate-gory) rather ilian moving to nearby plants Most ofthe beetles iliat we did recover were on plants inthe pots where they were released (Fig 4 dis-persal category none) Fifteen percent or less ofthe beetles dispersed to oilier plants in ilieir re-lease plot (Fig 4 within category) Five percentor less of beetles were recovered in nearby plots(Fig 4 other category) Flown beetles weremore likely to disperse locally in comparison withstored beetles about twice as many flown beetleswere recovered on other plants (within category)in the release plot or in nearby plots (oilier cat-egory) (Fig 4) in comparison witll stored beetles

Marking beetles wiili lacquer to distinguishtreatments had no apparent effect on aphid con-

B

69

c

A

Reared

c=J Flown

- Stored

12 24 36 48

Hours After Release

o

20

10

10

20

o

0 0aCD 20c-ooOJ

~ampl 10-w(f)

-tIX 0

Results

Beetle Dispersal Most adult H convergens re-leased on groups of aphid-infested potted plants(Fig 1) dispersed within 1-3 d (Figs 2 and 3) In2 of 5 trials (Figs 2C and 3A) allowing these ag-gregation-collected beetles to drink and fly (flown)in a screen tent for 7-10 d before release delayeddispersal from plots in comparison with beetlesthat were stored at 4-lOoC (stored) until the dayof release About 20 (SE = 7) and 36 (SE =13) more flown beetles were recovered in com-parison with stored beetles during September 1993(Fig 2C) (t = 228 n = 36 P = 0029) and July1994 (Fig 3A) (t = 260 n = 90 P = 0011)respectively when samples from all postreleasetimes (12-69 h) were combined During July 1993significantly more stored than flown beetles wererecovered during all sample times pooled (Fig 2B)(t = 303 n = 36 P lt 0005)

Insectary-reared beetles dispersed from plotsmore slowly than did aggregation-collected beetles(Fig 2A)About 19 (SE = 10) and 41 (SE =9) more reared beetles were recovered in compar-ison wiili flown (t = 209 n = 30 P = 0046) orstored (t = 389 n = 30 P = 00005) beetlesrespectively when all postrelease samples werepooled during May 1993 (Fig 2A)

Because treatments were marked wiili a differ-ent color and released into separate plots during1994 we were able to determine that most of ilie

arately for each treatment during both 1994 trials(July and September) The result is 8 observationsfor the regression at 72 h There were 5 observa-tions for the regression at 36 h because aphidswere not sampled at that interval during 1993

Percentage control (C) was calculated for eachtreatment according to the Abbott (1925) method

C = [(N - R)N] X 100

where N is the percentage of aphids alive (postre-lease density divided by prerelease density multi-plied by 100) on nonrelease plants and R is thepercentage of aphids alive on release plants at 36or 72 h after release

Statistical Analyses For each treatment wecompared mean aphid densities before and afterreleases and mean numbers of beetles recoveredat all postrelease times combined Beetles in non-release plots were excluded from the data andanalyses Less than 1 of beetles recovered duringeach trial were collected in nonrelease plots Un-less stated otherwise we used t-tests for 2-waycomparisons and analysis of variance (PROCANOVA SAS Institute 1988) for multiple compar-isons Because the beetle dispersal out categoryequals the sum of all other categories subtractedfrom tlle total number of beetles released allmeans were not independent and multiple t-testsinstead of ANOVA were used to compare beetledispersal

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 693

Fig 3 Mean + SEM number of adult convergentlady beetles recovered after a single release on chrysan-themums infested with melon aphids for trials conductedin July (A) and September (B) 1994 Beetles were col-lected from aggregations and were cold-stored until theday of release (stored) cold-stored until release and notlacquered (unmarked) or conditioned (flown)

sumption (see below) or dispersal During Septem-ber 1994 we released cold-stored H convergensthat were both marked (stored) and not marked(unmarked) and there were no obvious differencesin average recovery at individual sample times(Fig 3B) or when all samples were pooled (t =068 n = 120 P = 050) No natural populationsof convergent lady beetles were observed on sur-rounding vegetation in September and presum-ably all unmarked beetles we recovered were thosewe released

Aphid Control Despite relatively rapid dispers-al from plots release of convergent lady beetlessignificantly reduced melon aphid densities on pot-ted chrysanthemums during all trials (Figs 5 and6) Within 3 d after a single release of 34-42 bee-tles per pot H convergens provided 25-84 con-trol as calculated according to the Abbott (1925)method Up to 54 control occurred 15 dafterrelease (Table 1)

Each adult H convergens consumed =25-170melon aphids per day (Fig 7) as estimated by di-viding the average reduction in aphids by the av-erage number of beetles present Samples collect-

A

c=J FloWn

- Storedc=J Unmarked

ILB

III eOther

B

A

c=J Flown

- Stored

Within

e e e f

d

Out

100

80-c~ 60Q)

gt8~ 40CIlQ) 20~Q)

CD-c 0Q)Vllaquol(I) 80Qj00 a~ 600

LiJCI) 40plusmn-IX

20

0None

ed 3 d after release reveal a strong positiveassociation between aphid density per leaf at thetime of release and daily aphid consumption perbeetle (AB) AB was higher on plants with a bigherinitial aphid density (Fig 7) Feeding rate and ini-tial aphid density were not signiRcantly correlatedat 15 d (r2 = 065 F = 56 df = 1 3 P = 01)aphids were sampled at this interval only during1994 and sample size was small

The extent of predation by poundlown beetles in com-parison with stored beetles could not be deter-mined during 1993 because equal numbers of bee-tles from all treatments were released on each potHowever poundlown and stored beetles were releasedinto separate plots during 1994 and both treat-ments significantly reduced aphid densities in com-parison with plots without beetles (Fig 6) Therewere no apparent differences in daily aphid con-sumption per beetle among treatments (poundlown

Dispersal

Fig 4 Mean + SEM dispersal (percentage of re-leased beetles recovered) for flownand stored convergentlady beetles for all sample times pooled over 3 d duringJuly (A) or September (B) 1994 In total 2400 beetleswere released in Julyand 2400 beetles (1200 each flownand stored) were released during September Nondis-persing beetles (none) were recovered in pots where theywere released Beetles that left the study area (out) werenot recovered Some beetles moved from the pot wherethey were released but remained within their release plot(within)The remaining beetles were recovered after dis-persing from release plants to other plots (other) in ourstudy area Within each month means that do not sharea common letter are significantlydifferent (P lt 005 t-tests)

6912 24 36 48Hours After Release

o

40

30

0 20abullQ) 10aCIlQ)Q)Q) 0enLiJ 40CI)Ix 30

20

10

0

694 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

c=J No Beetles - With Beetles A

0250 c==J No Beetles Stored B

- Flown ~ Unmarked200

150

100

50

0Before 15 Days 3 Days

500

-lt1l3 100

300

400

200

bulllt1lC()s1sclaquoWC)+-Ix

B

c

A

80

250200150100

50

- 0~ 50JQ 40a~ 30s~ 20gr 10+I 0

120

Fig 5 Mean + SEM melon aphids per leaf beforeand 3 d after release outdoors of 34-42 adult lady beetlesper pot each pot containing 4 chrysanthemum plants fortrials conducted in May (A) July (B) and September (C)1993 Aphid means in each trial were not significantlydifferent (P gt 001) before release (A) t = 139 n = 1272 (B) t = 031 n = 48 72 (C) t = 008 n = 64 96Means were significantly different (P lt 00001) after re-lease (A) t = 468 n = 20 72 (B) t = 1186 n = 4872 (C) t = 574 n = 6496 Mean plusmn SEM leaves perplant (L) and plant height (H) in centimeters for all treat-ments pooled were (A) L = 154 (04) H = 287 (O(l)(B) L = 158 (04) H = 301 (08) (C) L = 171 (06)H = 297 (11)

40

oBefore Release After Release

Fig 6 Mean + SEM melon aphids per leaf beforeand 15 and 3 d after release outdoors of 40 lown storedor unmarked adult convergent lady beetles per pot eachpot containing 4 chrysanthemum plants for trials con-ducted in July (A) and September (B) 1994 Aphid meansin each trial were not significantly different (P gt 005)before release (A) F = 00 df = 2 213 (B) F = 002df = 3 187 There were significant differences (P lt00001) in aphid densities after release of convergent ladybeetles ANOVA statistics are (A) 15 d F = 195 df =2 69 3 d F = 412 df = 2 141 (B) 15 d F = 146df = 3 92 3 d F = 754 df = 3 116 Mean plusmn SEMleaves per plant (L) and plant height (H) in centimetersfor all treatments pooled were (A) L = 111 (04) H =176 (03) (B) L = 178 (04) H = 252 (04)

Table 1 Melon aphid control provided 15 and 3 dafter a single releasc of 34-42 convergent lady beetlcHper chrysanthemum pot

stored unmarked) (Fig 6B) However at 15 d af-ter release during July 1994 flown beetles provid-ed more control in comparison with stored beetles(Fig 6A) because flown beetles dispersed lessquickly (Fig 3A) so that more individuals werepresent to feed

Discussion

We found that preconditioning aggregation-col-lected H convergens or releasing reared beetles insome instances delayed dispersal in comparisonwith cold-stored beetles from aggregations Ourfindings differ from those of Starks et al (1975)who found no differences in dispersal after release

Mean SE controlMonth Year Treatmenti

15 d 3d

May 1993 All NS 345 (201)July 1993 All NS 784 43)Sept 1993 All NS 844 (59)July 1994 Flown 544 (111) 520 (64)Sept 1994 Flown 289 (147) 686 (63)July 1994 Stored 67 (160) 249 (129)Sept 1994 Stored 353 (116) 636 (58)Sept 1994 Unmarked 539 ( 70) 697 (69)

NS not sampled Calculated according to the Abbott (1925) methodb All all treatments (Hawn stored and reared) were released

on the same plants during 1993 so aphid consumption cannot beseparated by treatment and results from all treatments are com-bined

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 695

40

bull

Fi~ 7 Daily aphid consumption per adult convergentlady wtle (AB aphids per beetle mean t SEM) versusinitial density of melon aphids per leaf (AL aphids perleaf mean t SEM) on potted chrysantl1emums Theequation for the line is AB = 427 (SEM = 123) + 025(SEM = 005) AL (r2 = 082 F = 272 df = 1 6 P =0(02)

density-dependent functional response in preda-tion by H convergens has been demonstrated in amodel by Gutierrez et aI (1981) and has been re-ported for other aphid-feeding lady beetles in-cluding Cocci nella septempunctata L and Har-nwnia axyridis (Pallas) (Hukusima and Ohwaki1972 Hodek 1973)

Hippodamia convergens consumed a mean of 91(SEM = 94) aphids per day which agrees withHodeks (1973) conclusion that the daily feedingrate of adult aphidophagous Coccinellidae usuallyamounts to =100 aphids Clausen (1916) reared IIconvergens in the laboratory on hop aphids Pho-radon humuli (Schrank) which at maturity areabout the same size (1-2 mm body length) as mel-on aphids (Palmer 1952) Clausen found that adultH convergens over their life time consumed =66(SEM = 49 range 17-203) hop aphids each dayClausen used nearly full-grown aphids whereasours were of varying age and included smalleraphids which may account for the 13 lower num-ber of aphids eaten in Clausens (1916) study incomparison with our data Hagen and Sluss (1966)studied H convergens consumption of anothersmall species the spotted alfalfa aphid Therioa-phis maculata (Buckton) Preovipositional adultbeetles consumed 112 (SEM = 21) spotted alfalfaaphids per day and each beetle ate an average of91 aphids per day over their lifetime

Differences in aphid control or dispersal amongtreatments might be more dramatic with differentprerelease treatment of beetles Ignoffo et al(1977) found that clipping the wings of II conver-gens and other entomophages before release re-duced dispersal and increased predation Beetlesin our shldy were active during their 7-10 d out-doors in a screen tent before release (the flowntreatment) beetles moved with daily changes inshade and when disturbed during watering How-ever beetles spent little time actually flying in thetent The main difference between our treatmentswas that flown beetles were exposed to ambientlight wind temperature and humidity for 7-10 dbefore release in comparison with stored beetleswhich were kept at 4-1OdegC and constant dark untilthe day of release

With the exception of 1 lady beetle egg clusterin May 1993 (when reared beetles were released)no reproduction by H convergens was observedGiven rapid dispersal and lack of reproduction anycontrol from a single release would likely be tem-porary Releases are useful for inundative controlbut apparently are not effective for inoculative bi-ological control

Further research is needed to investigate the re-lationships among aphid density beetle releaserates and frequency dispersal and control Ourshldy plants harbored relatively high aphid densi-ties (Figs 5 and 6) and dispersal and control mightdiffer at the lower aphid densities more character-istic of a commercial nursery Hagen (1974) re-ports that the tendency of H convergens to feed

500100 200 300 400AphidsLeaf

oo

in sorghum of nondiapausing native H convergensin comparison with dispersal of aggregation-col-lected beetles Although preconditioning in ourstudy slowed dispersal treatments appear to havelittle practical benefit given that release of bothtreated and untreated beetles controlled aphids H(C)1lVlrglnsproduces few or no eggs unless fedaphids so rearing is expensive Preconditioning ina flight tent requires watering beetles and recol-lecting them before release Some beetles die dur-ing preconditioning reducing the number avail-able at the time of release Because H convergensis relatively inexpensive when purchased in bulk(=$40 buys 70000-80000 beetles) it is less ex-pensive to release greater numbers of stored bee-tles or to introduce them more often to compen-sate for their more rapid dispersal in comparisonwith preconditioning or rearing beetles

Although Starks et al (1975) found no differ-ences in dispersal among treatments native beetlesprovided better aphid control than aggregation-collected beetles When Starks et aI (1975) re-leased 2 beetles per plant in the greenhouse oncaged sorghum infested with 100-400 greenbugsSchizaphis graminum (Rondani) per plant theyobserved =95 and 67 aphid reduction respec-tively 5 d after release of native and aggregation-collected beetles These differences were appar-ently because beetles from aggregations oftenrested at the top of cages without feeding onaphids

Hippodamia convergens in our study consumed=25-170 melon aphids per day increasing con-sumption as aphid density increased (Fig 7) This

160 bull 19930 1994

Q) 120

fQ)Q)al-()C 80Ea4

696 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

aggregate and disperse varies with seasonal differ-ences in physiological condition of the beetlesHost plant aphid species release environmentand environmental conditions also may affectaphid control from beetle release

We are not aware of any studies indicating thatconvergent lady beetle releases are effective incontrolling aphids on a large scale Our preliminaryresearch on landscape roses indicates that releaseof relatively large numbers of beetles or multiplereleases may be needed to obtain control on largerplants Little is known about overwintering H con-vergens and the potential ecological effects oflarge-scale harvesting of aggregating lady beetleshave not been investigated Even if releases of ag-gregation-collected convergent lady beetles couldcontrol aphids on a large scale potential ecologicalor nontarget effects might make large-scale releas-es undesirable

There are quarantine considerations regardingthe shipment of field-collected insects Aggregat-ing beetles may harbor microorganisms and bulkcollections may be contaminated with other inver-tebrates The parasite Dinocampus (=Perilitus)coccinellae (Shrank) (Hymenoptera Braconidae)emerged from a few of the commercial H conver-gens we received We did not dissect hosts to checkparasitism and did not hold beetles and feed themaphids to induce D coccinellae emergence so wedo not know the extent of their parasitization byD coccinellae which attacks gt40 coccinellid spe-cies (Obrycki 1989) Ruzicka and Hagen (1985)found that 10 of the H convergens collectedfrom aggregations in California are parasitized byD coccinellae and parasitized beetles have less ofa tendency to disperse in comparison with unpar-asitized H convergens

Augmentative release of commercially availableconvergent lady beetles can provide inundativecontrol of relatively high aphid densities on sIllallpotted plants in a limited area Convergent ladybeetles are organically acceptable readily avail-able hardy easily handled and store well Pesti-cide exposure hazards and worker reentry restric-tions after pesticide application may make ladybeetle releases an attractive option for aphid con-trol in nurseries and greenhouses However ladybeetle releases must be integrated with other man-agement practices and pesticide applications forother pests may need to be modified for releasesto be effective

Acknowledgments

Michael P Parrella Norman E Gary Jill RentnerChuck LeGault and Bob Rousseau (Department of En-tomology University of California Davis) and Robin Ro-setta Pat Thompson and Ellen Zagory (Davis Arbore-tum University of California Davis) assisted usKenneth S Hagen and Kent M Daane (Laboratory ofBiological Control University of California Berkeley)Lester E Ehler (Department of Entomology Universityof California Davis) and Michael J Raupp (Department

of Entomology University of Maryland) critiqued draftsof the manuscript Unique Insect Control CitrusHeights CA provided the lady beetles The Universityof California Elvenia J Slosson Fund for OrnamentalHorticulture provided financial assistance

References Cited

Abbott W S 1925 A method of computing the ef-fectiveness of an insecticide J Econ Entomol 18265-267

Clausen C P 1916 Life-history and feeding recordsof a series of California Coccinellidae Univ CalifPub Entomol 1(6) 251-299

Cooke W C 1963 Ecology of the pea aphid in theblue mountain area of eastern Oregon and Washing-ton US Dep Agric Tech Bull 1287

Davidson W M 1919 The convergent ladybirdbeetle (Hippodamia convergens Guerin) and the bar-ley-com aphis (Aphis lIuzidis Fitch) Mon Bull CalifState Comm Homc 8 23-26

1924 Observations and experiments on the dispersionof the convergent lady-beetle (Hippodamia conver-gens Guerin-MenevilIe) in California Trans Am En-tomol Soc 50 163--175

Davis J R and R L Kirkland 1982 Physiologicaland environmental factors related to the dispersalflight of the convergent lady beetle Hippodamia con-vergens (Guerin-MenevilIe) J Kans Entomo Soc55 187-196

Eddy C O 1939 An attempt to colonize Hippodamiaconvergens Proc Int Soc Sugar Cane Technol 6385--386 (abstracted in Rev App Entomo 28 245[1940])

Fenton F A and R G Dahms 1951 Attempts atcontrolling the greenbug by the importation and re-lease of lady beetles in Oklahoma Proc Okla AcadSci 32 49-51

Flint M L S H Dreistadt J Rentner and M PParrella 1995 Lady beetle release controls aphidson potted plants Calif Agric 49(2) 5--8

Gutierrez A P J U Baumgaertner and K S Ha-gen 1981 A conceptual model for growth devel-opment and reproduction in the ladybird beetle Hip-podamia convergens (Coleoptera Coccinellidae)Can Entomo 113 21-33

Hagen K S 1962 Biology and ecology of predaceousCoccinellidae Annu Rev Entomol 7 289-326

1974 The significance of predaceous Coccinellidae inbiological and integrated control of insects Entomo-phaga 7 25-44

1982 Notes on the convergent lady beetle Univ CalifDiv Agric Nat Resour Leaf 2502

Hagen K S and R R Sluss 1966 Quantity ofaphids required for reproduction by Hippodamia sppin the laboratory pp 47-59 In I Hodek [ed] Ecol-ogy of aphidophagous insects Junk The Hague

Hatch M H and C Tanasse 1948 The liberationof Hippodamia convergens in the Yakima Valley ofWashington 1943 to 1946 J Econ EntomoI 41993

Hodek I 1973 Biology of Coccinellidae Junk TheHague

Hukusima S and T Ohwaki 1972 Further noteson feeding biology of Hannonia axyridis Pallas (Co-leoptera Coccinellidae) Res Bull Fac Agric GifuUniv 33 75--82

692 ENVIHONMENTAL ENTOMOLOGY Vol 25 no 3

Fig 2 Mean + SEM number of adult lady beetlesrecoveredafter a single release on chrysanlliemumsin-festedwillimelonaphidsfor trialsconductedin May(A)July (B) and September (e) 1993 Beetles were insec-tary-reared (reared) collected from aggregationsandcold-storeduntil the day of release (stored)or collectedfrom aggregationsand conditioned(flown)

beetles we recovered were in the same plots wherethey were released Because we also resampledplants from which we had previously removed allbeetles we found that when beetles did dispersethey usually left our study site (Fig 4 out cate-gory) rather ilian moving to nearby plants Most ofthe beetles iliat we did recover were on plants inthe pots where they were released (Fig 4 dis-persal category none) Fifteen percent or less ofthe beetles dispersed to oilier plants in ilieir re-lease plot (Fig 4 within category) Five percentor less of beetles were recovered in nearby plots(Fig 4 other category) Flown beetles weremore likely to disperse locally in comparison withstored beetles about twice as many flown beetleswere recovered on other plants (within category)in the release plot or in nearby plots (oilier cat-egory) (Fig 4) in comparison witll stored beetles

Marking beetles wiili lacquer to distinguishtreatments had no apparent effect on aphid con-

B

69

c

A

Reared

c=J Flown

- Stored

12 24 36 48

Hours After Release

o

20

10

10

20

o

0 0aCD 20c-ooOJ

~ampl 10-w(f)

-tIX 0

Results

Beetle Dispersal Most adult H convergens re-leased on groups of aphid-infested potted plants(Fig 1) dispersed within 1-3 d (Figs 2 and 3) In2 of 5 trials (Figs 2C and 3A) allowing these ag-gregation-collected beetles to drink and fly (flown)in a screen tent for 7-10 d before release delayeddispersal from plots in comparison with beetlesthat were stored at 4-lOoC (stored) until the dayof release About 20 (SE = 7) and 36 (SE =13) more flown beetles were recovered in com-parison with stored beetles during September 1993(Fig 2C) (t = 228 n = 36 P = 0029) and July1994 (Fig 3A) (t = 260 n = 90 P = 0011)respectively when samples from all postreleasetimes (12-69 h) were combined During July 1993significantly more stored than flown beetles wererecovered during all sample times pooled (Fig 2B)(t = 303 n = 36 P lt 0005)

Insectary-reared beetles dispersed from plotsmore slowly than did aggregation-collected beetles(Fig 2A)About 19 (SE = 10) and 41 (SE =9) more reared beetles were recovered in compar-ison wiili flown (t = 209 n = 30 P = 0046) orstored (t = 389 n = 30 P = 00005) beetlesrespectively when all postrelease samples werepooled during May 1993 (Fig 2A)

Because treatments were marked wiili a differ-ent color and released into separate plots during1994 we were able to determine that most of ilie

arately for each treatment during both 1994 trials(July and September) The result is 8 observationsfor the regression at 72 h There were 5 observa-tions for the regression at 36 h because aphidswere not sampled at that interval during 1993

Percentage control (C) was calculated for eachtreatment according to the Abbott (1925) method

C = [(N - R)N] X 100

where N is the percentage of aphids alive (postre-lease density divided by prerelease density multi-plied by 100) on nonrelease plants and R is thepercentage of aphids alive on release plants at 36or 72 h after release

Statistical Analyses For each treatment wecompared mean aphid densities before and afterreleases and mean numbers of beetles recoveredat all postrelease times combined Beetles in non-release plots were excluded from the data andanalyses Less than 1 of beetles recovered duringeach trial were collected in nonrelease plots Un-less stated otherwise we used t-tests for 2-waycomparisons and analysis of variance (PROCANOVA SAS Institute 1988) for multiple compar-isons Because the beetle dispersal out categoryequals the sum of all other categories subtractedfrom tlle total number of beetles released allmeans were not independent and multiple t-testsinstead of ANOVA were used to compare beetledispersal

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 693

Fig 3 Mean + SEM number of adult convergentlady beetles recovered after a single release on chrysan-themums infested with melon aphids for trials conductedin July (A) and September (B) 1994 Beetles were col-lected from aggregations and were cold-stored until theday of release (stored) cold-stored until release and notlacquered (unmarked) or conditioned (flown)

sumption (see below) or dispersal During Septem-ber 1994 we released cold-stored H convergensthat were both marked (stored) and not marked(unmarked) and there were no obvious differencesin average recovery at individual sample times(Fig 3B) or when all samples were pooled (t =068 n = 120 P = 050) No natural populationsof convergent lady beetles were observed on sur-rounding vegetation in September and presum-ably all unmarked beetles we recovered were thosewe released

Aphid Control Despite relatively rapid dispers-al from plots release of convergent lady beetlessignificantly reduced melon aphid densities on pot-ted chrysanthemums during all trials (Figs 5 and6) Within 3 d after a single release of 34-42 bee-tles per pot H convergens provided 25-84 con-trol as calculated according to the Abbott (1925)method Up to 54 control occurred 15 dafterrelease (Table 1)

Each adult H convergens consumed =25-170melon aphids per day (Fig 7) as estimated by di-viding the average reduction in aphids by the av-erage number of beetles present Samples collect-

A

c=J FloWn

- Storedc=J Unmarked

ILB

III eOther

B

A

c=J Flown

- Stored

Within

e e e f

d

Out

100

80-c~ 60Q)

gt8~ 40CIlQ) 20~Q)

CD-c 0Q)Vllaquol(I) 80Qj00 a~ 600

LiJCI) 40plusmn-IX

20

0None

ed 3 d after release reveal a strong positiveassociation between aphid density per leaf at thetime of release and daily aphid consumption perbeetle (AB) AB was higher on plants with a bigherinitial aphid density (Fig 7) Feeding rate and ini-tial aphid density were not signiRcantly correlatedat 15 d (r2 = 065 F = 56 df = 1 3 P = 01)aphids were sampled at this interval only during1994 and sample size was small

The extent of predation by poundlown beetles in com-parison with stored beetles could not be deter-mined during 1993 because equal numbers of bee-tles from all treatments were released on each potHowever poundlown and stored beetles were releasedinto separate plots during 1994 and both treat-ments significantly reduced aphid densities in com-parison with plots without beetles (Fig 6) Therewere no apparent differences in daily aphid con-sumption per beetle among treatments (poundlown

Dispersal

Fig 4 Mean + SEM dispersal (percentage of re-leased beetles recovered) for flownand stored convergentlady beetles for all sample times pooled over 3 d duringJuly (A) or September (B) 1994 In total 2400 beetleswere released in Julyand 2400 beetles (1200 each flownand stored) were released during September Nondis-persing beetles (none) were recovered in pots where theywere released Beetles that left the study area (out) werenot recovered Some beetles moved from the pot wherethey were released but remained within their release plot(within)The remaining beetles were recovered after dis-persing from release plants to other plots (other) in ourstudy area Within each month means that do not sharea common letter are significantlydifferent (P lt 005 t-tests)

6912 24 36 48Hours After Release

o

40

30

0 20abullQ) 10aCIlQ)Q)Q) 0enLiJ 40CI)Ix 30

20

10

0

694 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

c=J No Beetles - With Beetles A

0250 c==J No Beetles Stored B

- Flown ~ Unmarked200

150

100

50

0Before 15 Days 3 Days

500

-lt1l3 100

300

400

200

bulllt1lC()s1sclaquoWC)+-Ix

B

c

A

80

250200150100

50

- 0~ 50JQ 40a~ 30s~ 20gr 10+I 0

120

Fig 5 Mean + SEM melon aphids per leaf beforeand 3 d after release outdoors of 34-42 adult lady beetlesper pot each pot containing 4 chrysanthemum plants fortrials conducted in May (A) July (B) and September (C)1993 Aphid means in each trial were not significantlydifferent (P gt 001) before release (A) t = 139 n = 1272 (B) t = 031 n = 48 72 (C) t = 008 n = 64 96Means were significantly different (P lt 00001) after re-lease (A) t = 468 n = 20 72 (B) t = 1186 n = 4872 (C) t = 574 n = 6496 Mean plusmn SEM leaves perplant (L) and plant height (H) in centimeters for all treat-ments pooled were (A) L = 154 (04) H = 287 (O(l)(B) L = 158 (04) H = 301 (08) (C) L = 171 (06)H = 297 (11)

40

oBefore Release After Release

Fig 6 Mean + SEM melon aphids per leaf beforeand 15 and 3 d after release outdoors of 40 lown storedor unmarked adult convergent lady beetles per pot eachpot containing 4 chrysanthemum plants for trials con-ducted in July (A) and September (B) 1994 Aphid meansin each trial were not significantly different (P gt 005)before release (A) F = 00 df = 2 213 (B) F = 002df = 3 187 There were significant differences (P lt00001) in aphid densities after release of convergent ladybeetles ANOVA statistics are (A) 15 d F = 195 df =2 69 3 d F = 412 df = 2 141 (B) 15 d F = 146df = 3 92 3 d F = 754 df = 3 116 Mean plusmn SEMleaves per plant (L) and plant height (H) in centimetersfor all treatments pooled were (A) L = 111 (04) H =176 (03) (B) L = 178 (04) H = 252 (04)

Table 1 Melon aphid control provided 15 and 3 dafter a single releasc of 34-42 convergent lady beetlcHper chrysanthemum pot

stored unmarked) (Fig 6B) However at 15 d af-ter release during July 1994 flown beetles provid-ed more control in comparison with stored beetles(Fig 6A) because flown beetles dispersed lessquickly (Fig 3A) so that more individuals werepresent to feed

Discussion

We found that preconditioning aggregation-col-lected H convergens or releasing reared beetles insome instances delayed dispersal in comparisonwith cold-stored beetles from aggregations Ourfindings differ from those of Starks et al (1975)who found no differences in dispersal after release

Mean SE controlMonth Year Treatmenti

15 d 3d

May 1993 All NS 345 (201)July 1993 All NS 784 43)Sept 1993 All NS 844 (59)July 1994 Flown 544 (111) 520 (64)Sept 1994 Flown 289 (147) 686 (63)July 1994 Stored 67 (160) 249 (129)Sept 1994 Stored 353 (116) 636 (58)Sept 1994 Unmarked 539 ( 70) 697 (69)

NS not sampled Calculated according to the Abbott (1925) methodb All all treatments (Hawn stored and reared) were released

on the same plants during 1993 so aphid consumption cannot beseparated by treatment and results from all treatments are com-bined

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 695

40

bull

Fi~ 7 Daily aphid consumption per adult convergentlady wtle (AB aphids per beetle mean t SEM) versusinitial density of melon aphids per leaf (AL aphids perleaf mean t SEM) on potted chrysantl1emums Theequation for the line is AB = 427 (SEM = 123) + 025(SEM = 005) AL (r2 = 082 F = 272 df = 1 6 P =0(02)

density-dependent functional response in preda-tion by H convergens has been demonstrated in amodel by Gutierrez et aI (1981) and has been re-ported for other aphid-feeding lady beetles in-cluding Cocci nella septempunctata L and Har-nwnia axyridis (Pallas) (Hukusima and Ohwaki1972 Hodek 1973)

Hippodamia convergens consumed a mean of 91(SEM = 94) aphids per day which agrees withHodeks (1973) conclusion that the daily feedingrate of adult aphidophagous Coccinellidae usuallyamounts to =100 aphids Clausen (1916) reared IIconvergens in the laboratory on hop aphids Pho-radon humuli (Schrank) which at maturity areabout the same size (1-2 mm body length) as mel-on aphids (Palmer 1952) Clausen found that adultH convergens over their life time consumed =66(SEM = 49 range 17-203) hop aphids each dayClausen used nearly full-grown aphids whereasours were of varying age and included smalleraphids which may account for the 13 lower num-ber of aphids eaten in Clausens (1916) study incomparison with our data Hagen and Sluss (1966)studied H convergens consumption of anothersmall species the spotted alfalfa aphid Therioa-phis maculata (Buckton) Preovipositional adultbeetles consumed 112 (SEM = 21) spotted alfalfaaphids per day and each beetle ate an average of91 aphids per day over their lifetime

Differences in aphid control or dispersal amongtreatments might be more dramatic with differentprerelease treatment of beetles Ignoffo et al(1977) found that clipping the wings of II conver-gens and other entomophages before release re-duced dispersal and increased predation Beetlesin our shldy were active during their 7-10 d out-doors in a screen tent before release (the flowntreatment) beetles moved with daily changes inshade and when disturbed during watering How-ever beetles spent little time actually flying in thetent The main difference between our treatmentswas that flown beetles were exposed to ambientlight wind temperature and humidity for 7-10 dbefore release in comparison with stored beetleswhich were kept at 4-1OdegC and constant dark untilthe day of release

With the exception of 1 lady beetle egg clusterin May 1993 (when reared beetles were released)no reproduction by H convergens was observedGiven rapid dispersal and lack of reproduction anycontrol from a single release would likely be tem-porary Releases are useful for inundative controlbut apparently are not effective for inoculative bi-ological control

Further research is needed to investigate the re-lationships among aphid density beetle releaserates and frequency dispersal and control Ourshldy plants harbored relatively high aphid densi-ties (Figs 5 and 6) and dispersal and control mightdiffer at the lower aphid densities more character-istic of a commercial nursery Hagen (1974) re-ports that the tendency of H convergens to feed

500100 200 300 400AphidsLeaf

oo

in sorghum of nondiapausing native H convergensin comparison with dispersal of aggregation-col-lected beetles Although preconditioning in ourstudy slowed dispersal treatments appear to havelittle practical benefit given that release of bothtreated and untreated beetles controlled aphids H(C)1lVlrglnsproduces few or no eggs unless fedaphids so rearing is expensive Preconditioning ina flight tent requires watering beetles and recol-lecting them before release Some beetles die dur-ing preconditioning reducing the number avail-able at the time of release Because H convergensis relatively inexpensive when purchased in bulk(=$40 buys 70000-80000 beetles) it is less ex-pensive to release greater numbers of stored bee-tles or to introduce them more often to compen-sate for their more rapid dispersal in comparisonwith preconditioning or rearing beetles

Although Starks et al (1975) found no differ-ences in dispersal among treatments native beetlesprovided better aphid control than aggregation-collected beetles When Starks et aI (1975) re-leased 2 beetles per plant in the greenhouse oncaged sorghum infested with 100-400 greenbugsSchizaphis graminum (Rondani) per plant theyobserved =95 and 67 aphid reduction respec-tively 5 d after release of native and aggregation-collected beetles These differences were appar-ently because beetles from aggregations oftenrested at the top of cages without feeding onaphids

Hippodamia convergens in our study consumed=25-170 melon aphids per day increasing con-sumption as aphid density increased (Fig 7) This

160 bull 19930 1994

Q) 120

fQ)Q)al-()C 80Ea4

696 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

aggregate and disperse varies with seasonal differ-ences in physiological condition of the beetlesHost plant aphid species release environmentand environmental conditions also may affectaphid control from beetle release

We are not aware of any studies indicating thatconvergent lady beetle releases are effective incontrolling aphids on a large scale Our preliminaryresearch on landscape roses indicates that releaseof relatively large numbers of beetles or multiplereleases may be needed to obtain control on largerplants Little is known about overwintering H con-vergens and the potential ecological effects oflarge-scale harvesting of aggregating lady beetleshave not been investigated Even if releases of ag-gregation-collected convergent lady beetles couldcontrol aphids on a large scale potential ecologicalor nontarget effects might make large-scale releas-es undesirable

There are quarantine considerations regardingthe shipment of field-collected insects Aggregat-ing beetles may harbor microorganisms and bulkcollections may be contaminated with other inver-tebrates The parasite Dinocampus (=Perilitus)coccinellae (Shrank) (Hymenoptera Braconidae)emerged from a few of the commercial H conver-gens we received We did not dissect hosts to checkparasitism and did not hold beetles and feed themaphids to induce D coccinellae emergence so wedo not know the extent of their parasitization byD coccinellae which attacks gt40 coccinellid spe-cies (Obrycki 1989) Ruzicka and Hagen (1985)found that 10 of the H convergens collectedfrom aggregations in California are parasitized byD coccinellae and parasitized beetles have less ofa tendency to disperse in comparison with unpar-asitized H convergens

Augmentative release of commercially availableconvergent lady beetles can provide inundativecontrol of relatively high aphid densities on sIllallpotted plants in a limited area Convergent ladybeetles are organically acceptable readily avail-able hardy easily handled and store well Pesti-cide exposure hazards and worker reentry restric-tions after pesticide application may make ladybeetle releases an attractive option for aphid con-trol in nurseries and greenhouses However ladybeetle releases must be integrated with other man-agement practices and pesticide applications forother pests may need to be modified for releasesto be effective

Acknowledgments

Michael P Parrella Norman E Gary Jill RentnerChuck LeGault and Bob Rousseau (Department of En-tomology University of California Davis) and Robin Ro-setta Pat Thompson and Ellen Zagory (Davis Arbore-tum University of California Davis) assisted usKenneth S Hagen and Kent M Daane (Laboratory ofBiological Control University of California Berkeley)Lester E Ehler (Department of Entomology Universityof California Davis) and Michael J Raupp (Department

of Entomology University of Maryland) critiqued draftsof the manuscript Unique Insect Control CitrusHeights CA provided the lady beetles The Universityof California Elvenia J Slosson Fund for OrnamentalHorticulture provided financial assistance

References Cited

Abbott W S 1925 A method of computing the ef-fectiveness of an insecticide J Econ Entomol 18265-267

Clausen C P 1916 Life-history and feeding recordsof a series of California Coccinellidae Univ CalifPub Entomol 1(6) 251-299

Cooke W C 1963 Ecology of the pea aphid in theblue mountain area of eastern Oregon and Washing-ton US Dep Agric Tech Bull 1287

Davidson W M 1919 The convergent ladybirdbeetle (Hippodamia convergens Guerin) and the bar-ley-com aphis (Aphis lIuzidis Fitch) Mon Bull CalifState Comm Homc 8 23-26

1924 Observations and experiments on the dispersionof the convergent lady-beetle (Hippodamia conver-gens Guerin-MenevilIe) in California Trans Am En-tomol Soc 50 163--175

Davis J R and R L Kirkland 1982 Physiologicaland environmental factors related to the dispersalflight of the convergent lady beetle Hippodamia con-vergens (Guerin-MenevilIe) J Kans Entomo Soc55 187-196

Eddy C O 1939 An attempt to colonize Hippodamiaconvergens Proc Int Soc Sugar Cane Technol 6385--386 (abstracted in Rev App Entomo 28 245[1940])

Fenton F A and R G Dahms 1951 Attempts atcontrolling the greenbug by the importation and re-lease of lady beetles in Oklahoma Proc Okla AcadSci 32 49-51

Flint M L S H Dreistadt J Rentner and M PParrella 1995 Lady beetle release controls aphidson potted plants Calif Agric 49(2) 5--8

Gutierrez A P J U Baumgaertner and K S Ha-gen 1981 A conceptual model for growth devel-opment and reproduction in the ladybird beetle Hip-podamia convergens (Coleoptera Coccinellidae)Can Entomo 113 21-33

Hagen K S 1962 Biology and ecology of predaceousCoccinellidae Annu Rev Entomol 7 289-326

1974 The significance of predaceous Coccinellidae inbiological and integrated control of insects Entomo-phaga 7 25-44

1982 Notes on the convergent lady beetle Univ CalifDiv Agric Nat Resour Leaf 2502

Hagen K S and R R Sluss 1966 Quantity ofaphids required for reproduction by Hippodamia sppin the laboratory pp 47-59 In I Hodek [ed] Ecol-ogy of aphidophagous insects Junk The Hague

Hatch M H and C Tanasse 1948 The liberationof Hippodamia convergens in the Yakima Valley ofWashington 1943 to 1946 J Econ EntomoI 41993

Hodek I 1973 Biology of Coccinellidae Junk TheHague

Hukusima S and T Ohwaki 1972 Further noteson feeding biology of Hannonia axyridis Pallas (Co-leoptera Coccinellidae) Res Bull Fac Agric GifuUniv 33 75--82

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 693

Fig 3 Mean + SEM number of adult convergentlady beetles recovered after a single release on chrysan-themums infested with melon aphids for trials conductedin July (A) and September (B) 1994 Beetles were col-lected from aggregations and were cold-stored until theday of release (stored) cold-stored until release and notlacquered (unmarked) or conditioned (flown)

sumption (see below) or dispersal During Septem-ber 1994 we released cold-stored H convergensthat were both marked (stored) and not marked(unmarked) and there were no obvious differencesin average recovery at individual sample times(Fig 3B) or when all samples were pooled (t =068 n = 120 P = 050) No natural populationsof convergent lady beetles were observed on sur-rounding vegetation in September and presum-ably all unmarked beetles we recovered were thosewe released

Aphid Control Despite relatively rapid dispers-al from plots release of convergent lady beetlessignificantly reduced melon aphid densities on pot-ted chrysanthemums during all trials (Figs 5 and6) Within 3 d after a single release of 34-42 bee-tles per pot H convergens provided 25-84 con-trol as calculated according to the Abbott (1925)method Up to 54 control occurred 15 dafterrelease (Table 1)

Each adult H convergens consumed =25-170melon aphids per day (Fig 7) as estimated by di-viding the average reduction in aphids by the av-erage number of beetles present Samples collect-

A

c=J FloWn

- Storedc=J Unmarked

ILB

III eOther

B

A

c=J Flown

- Stored

Within

e e e f

d

Out

100

80-c~ 60Q)

gt8~ 40CIlQ) 20~Q)

CD-c 0Q)Vllaquol(I) 80Qj00 a~ 600

LiJCI) 40plusmn-IX

20

0None

ed 3 d after release reveal a strong positiveassociation between aphid density per leaf at thetime of release and daily aphid consumption perbeetle (AB) AB was higher on plants with a bigherinitial aphid density (Fig 7) Feeding rate and ini-tial aphid density were not signiRcantly correlatedat 15 d (r2 = 065 F = 56 df = 1 3 P = 01)aphids were sampled at this interval only during1994 and sample size was small

The extent of predation by poundlown beetles in com-parison with stored beetles could not be deter-mined during 1993 because equal numbers of bee-tles from all treatments were released on each potHowever poundlown and stored beetles were releasedinto separate plots during 1994 and both treat-ments significantly reduced aphid densities in com-parison with plots without beetles (Fig 6) Therewere no apparent differences in daily aphid con-sumption per beetle among treatments (poundlown

Dispersal

Fig 4 Mean + SEM dispersal (percentage of re-leased beetles recovered) for flownand stored convergentlady beetles for all sample times pooled over 3 d duringJuly (A) or September (B) 1994 In total 2400 beetleswere released in Julyand 2400 beetles (1200 each flownand stored) were released during September Nondis-persing beetles (none) were recovered in pots where theywere released Beetles that left the study area (out) werenot recovered Some beetles moved from the pot wherethey were released but remained within their release plot(within)The remaining beetles were recovered after dis-persing from release plants to other plots (other) in ourstudy area Within each month means that do not sharea common letter are significantlydifferent (P lt 005 t-tests)

6912 24 36 48Hours After Release

o

40

30

0 20abullQ) 10aCIlQ)Q)Q) 0enLiJ 40CI)Ix 30

20

10

0

694 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

c=J No Beetles - With Beetles A

0250 c==J No Beetles Stored B

- Flown ~ Unmarked200

150

100

50

0Before 15 Days 3 Days

500

-lt1l3 100

300

400

200

bulllt1lC()s1sclaquoWC)+-Ix

B

c

A

80

250200150100

50

- 0~ 50JQ 40a~ 30s~ 20gr 10+I 0

120

Fig 5 Mean + SEM melon aphids per leaf beforeand 3 d after release outdoors of 34-42 adult lady beetlesper pot each pot containing 4 chrysanthemum plants fortrials conducted in May (A) July (B) and September (C)1993 Aphid means in each trial were not significantlydifferent (P gt 001) before release (A) t = 139 n = 1272 (B) t = 031 n = 48 72 (C) t = 008 n = 64 96Means were significantly different (P lt 00001) after re-lease (A) t = 468 n = 20 72 (B) t = 1186 n = 4872 (C) t = 574 n = 6496 Mean plusmn SEM leaves perplant (L) and plant height (H) in centimeters for all treat-ments pooled were (A) L = 154 (04) H = 287 (O(l)(B) L = 158 (04) H = 301 (08) (C) L = 171 (06)H = 297 (11)

40

oBefore Release After Release

Fig 6 Mean + SEM melon aphids per leaf beforeand 15 and 3 d after release outdoors of 40 lown storedor unmarked adult convergent lady beetles per pot eachpot containing 4 chrysanthemum plants for trials con-ducted in July (A) and September (B) 1994 Aphid meansin each trial were not significantly different (P gt 005)before release (A) F = 00 df = 2 213 (B) F = 002df = 3 187 There were significant differences (P lt00001) in aphid densities after release of convergent ladybeetles ANOVA statistics are (A) 15 d F = 195 df =2 69 3 d F = 412 df = 2 141 (B) 15 d F = 146df = 3 92 3 d F = 754 df = 3 116 Mean plusmn SEMleaves per plant (L) and plant height (H) in centimetersfor all treatments pooled were (A) L = 111 (04) H =176 (03) (B) L = 178 (04) H = 252 (04)

Table 1 Melon aphid control provided 15 and 3 dafter a single releasc of 34-42 convergent lady beetlcHper chrysanthemum pot

stored unmarked) (Fig 6B) However at 15 d af-ter release during July 1994 flown beetles provid-ed more control in comparison with stored beetles(Fig 6A) because flown beetles dispersed lessquickly (Fig 3A) so that more individuals werepresent to feed

Discussion

We found that preconditioning aggregation-col-lected H convergens or releasing reared beetles insome instances delayed dispersal in comparisonwith cold-stored beetles from aggregations Ourfindings differ from those of Starks et al (1975)who found no differences in dispersal after release

Mean SE controlMonth Year Treatmenti

15 d 3d

May 1993 All NS 345 (201)July 1993 All NS 784 43)Sept 1993 All NS 844 (59)July 1994 Flown 544 (111) 520 (64)Sept 1994 Flown 289 (147) 686 (63)July 1994 Stored 67 (160) 249 (129)Sept 1994 Stored 353 (116) 636 (58)Sept 1994 Unmarked 539 ( 70) 697 (69)

NS not sampled Calculated according to the Abbott (1925) methodb All all treatments (Hawn stored and reared) were released

on the same plants during 1993 so aphid consumption cannot beseparated by treatment and results from all treatments are com-bined

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 695

40

bull

Fi~ 7 Daily aphid consumption per adult convergentlady wtle (AB aphids per beetle mean t SEM) versusinitial density of melon aphids per leaf (AL aphids perleaf mean t SEM) on potted chrysantl1emums Theequation for the line is AB = 427 (SEM = 123) + 025(SEM = 005) AL (r2 = 082 F = 272 df = 1 6 P =0(02)

density-dependent functional response in preda-tion by H convergens has been demonstrated in amodel by Gutierrez et aI (1981) and has been re-ported for other aphid-feeding lady beetles in-cluding Cocci nella septempunctata L and Har-nwnia axyridis (Pallas) (Hukusima and Ohwaki1972 Hodek 1973)

Hippodamia convergens consumed a mean of 91(SEM = 94) aphids per day which agrees withHodeks (1973) conclusion that the daily feedingrate of adult aphidophagous Coccinellidae usuallyamounts to =100 aphids Clausen (1916) reared IIconvergens in the laboratory on hop aphids Pho-radon humuli (Schrank) which at maturity areabout the same size (1-2 mm body length) as mel-on aphids (Palmer 1952) Clausen found that adultH convergens over their life time consumed =66(SEM = 49 range 17-203) hop aphids each dayClausen used nearly full-grown aphids whereasours were of varying age and included smalleraphids which may account for the 13 lower num-ber of aphids eaten in Clausens (1916) study incomparison with our data Hagen and Sluss (1966)studied H convergens consumption of anothersmall species the spotted alfalfa aphid Therioa-phis maculata (Buckton) Preovipositional adultbeetles consumed 112 (SEM = 21) spotted alfalfaaphids per day and each beetle ate an average of91 aphids per day over their lifetime

Differences in aphid control or dispersal amongtreatments might be more dramatic with differentprerelease treatment of beetles Ignoffo et al(1977) found that clipping the wings of II conver-gens and other entomophages before release re-duced dispersal and increased predation Beetlesin our shldy were active during their 7-10 d out-doors in a screen tent before release (the flowntreatment) beetles moved with daily changes inshade and when disturbed during watering How-ever beetles spent little time actually flying in thetent The main difference between our treatmentswas that flown beetles were exposed to ambientlight wind temperature and humidity for 7-10 dbefore release in comparison with stored beetleswhich were kept at 4-1OdegC and constant dark untilthe day of release

With the exception of 1 lady beetle egg clusterin May 1993 (when reared beetles were released)no reproduction by H convergens was observedGiven rapid dispersal and lack of reproduction anycontrol from a single release would likely be tem-porary Releases are useful for inundative controlbut apparently are not effective for inoculative bi-ological control

Further research is needed to investigate the re-lationships among aphid density beetle releaserates and frequency dispersal and control Ourshldy plants harbored relatively high aphid densi-ties (Figs 5 and 6) and dispersal and control mightdiffer at the lower aphid densities more character-istic of a commercial nursery Hagen (1974) re-ports that the tendency of H convergens to feed

500100 200 300 400AphidsLeaf

oo

in sorghum of nondiapausing native H convergensin comparison with dispersal of aggregation-col-lected beetles Although preconditioning in ourstudy slowed dispersal treatments appear to havelittle practical benefit given that release of bothtreated and untreated beetles controlled aphids H(C)1lVlrglnsproduces few or no eggs unless fedaphids so rearing is expensive Preconditioning ina flight tent requires watering beetles and recol-lecting them before release Some beetles die dur-ing preconditioning reducing the number avail-able at the time of release Because H convergensis relatively inexpensive when purchased in bulk(=$40 buys 70000-80000 beetles) it is less ex-pensive to release greater numbers of stored bee-tles or to introduce them more often to compen-sate for their more rapid dispersal in comparisonwith preconditioning or rearing beetles

Although Starks et al (1975) found no differ-ences in dispersal among treatments native beetlesprovided better aphid control than aggregation-collected beetles When Starks et aI (1975) re-leased 2 beetles per plant in the greenhouse oncaged sorghum infested with 100-400 greenbugsSchizaphis graminum (Rondani) per plant theyobserved =95 and 67 aphid reduction respec-tively 5 d after release of native and aggregation-collected beetles These differences were appar-ently because beetles from aggregations oftenrested at the top of cages without feeding onaphids

Hippodamia convergens in our study consumed=25-170 melon aphids per day increasing con-sumption as aphid density increased (Fig 7) This

160 bull 19930 1994

Q) 120

fQ)Q)al-()C 80Ea4

696 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

aggregate and disperse varies with seasonal differ-ences in physiological condition of the beetlesHost plant aphid species release environmentand environmental conditions also may affectaphid control from beetle release

We are not aware of any studies indicating thatconvergent lady beetle releases are effective incontrolling aphids on a large scale Our preliminaryresearch on landscape roses indicates that releaseof relatively large numbers of beetles or multiplereleases may be needed to obtain control on largerplants Little is known about overwintering H con-vergens and the potential ecological effects oflarge-scale harvesting of aggregating lady beetleshave not been investigated Even if releases of ag-gregation-collected convergent lady beetles couldcontrol aphids on a large scale potential ecologicalor nontarget effects might make large-scale releas-es undesirable

There are quarantine considerations regardingthe shipment of field-collected insects Aggregat-ing beetles may harbor microorganisms and bulkcollections may be contaminated with other inver-tebrates The parasite Dinocampus (=Perilitus)coccinellae (Shrank) (Hymenoptera Braconidae)emerged from a few of the commercial H conver-gens we received We did not dissect hosts to checkparasitism and did not hold beetles and feed themaphids to induce D coccinellae emergence so wedo not know the extent of their parasitization byD coccinellae which attacks gt40 coccinellid spe-cies (Obrycki 1989) Ruzicka and Hagen (1985)found that 10 of the H convergens collectedfrom aggregations in California are parasitized byD coccinellae and parasitized beetles have less ofa tendency to disperse in comparison with unpar-asitized H convergens

Augmentative release of commercially availableconvergent lady beetles can provide inundativecontrol of relatively high aphid densities on sIllallpotted plants in a limited area Convergent ladybeetles are organically acceptable readily avail-able hardy easily handled and store well Pesti-cide exposure hazards and worker reentry restric-tions after pesticide application may make ladybeetle releases an attractive option for aphid con-trol in nurseries and greenhouses However ladybeetle releases must be integrated with other man-agement practices and pesticide applications forother pests may need to be modified for releasesto be effective

Acknowledgments

Michael P Parrella Norman E Gary Jill RentnerChuck LeGault and Bob Rousseau (Department of En-tomology University of California Davis) and Robin Ro-setta Pat Thompson and Ellen Zagory (Davis Arbore-tum University of California Davis) assisted usKenneth S Hagen and Kent M Daane (Laboratory ofBiological Control University of California Berkeley)Lester E Ehler (Department of Entomology Universityof California Davis) and Michael J Raupp (Department

of Entomology University of Maryland) critiqued draftsof the manuscript Unique Insect Control CitrusHeights CA provided the lady beetles The Universityof California Elvenia J Slosson Fund for OrnamentalHorticulture provided financial assistance

References Cited

Abbott W S 1925 A method of computing the ef-fectiveness of an insecticide J Econ Entomol 18265-267

Clausen C P 1916 Life-history and feeding recordsof a series of California Coccinellidae Univ CalifPub Entomol 1(6) 251-299

Cooke W C 1963 Ecology of the pea aphid in theblue mountain area of eastern Oregon and Washing-ton US Dep Agric Tech Bull 1287

Davidson W M 1919 The convergent ladybirdbeetle (Hippodamia convergens Guerin) and the bar-ley-com aphis (Aphis lIuzidis Fitch) Mon Bull CalifState Comm Homc 8 23-26

1924 Observations and experiments on the dispersionof the convergent lady-beetle (Hippodamia conver-gens Guerin-MenevilIe) in California Trans Am En-tomol Soc 50 163--175

Davis J R and R L Kirkland 1982 Physiologicaland environmental factors related to the dispersalflight of the convergent lady beetle Hippodamia con-vergens (Guerin-MenevilIe) J Kans Entomo Soc55 187-196

Eddy C O 1939 An attempt to colonize Hippodamiaconvergens Proc Int Soc Sugar Cane Technol 6385--386 (abstracted in Rev App Entomo 28 245[1940])

Fenton F A and R G Dahms 1951 Attempts atcontrolling the greenbug by the importation and re-lease of lady beetles in Oklahoma Proc Okla AcadSci 32 49-51

Flint M L S H Dreistadt J Rentner and M PParrella 1995 Lady beetle release controls aphidson potted plants Calif Agric 49(2) 5--8

Gutierrez A P J U Baumgaertner and K S Ha-gen 1981 A conceptual model for growth devel-opment and reproduction in the ladybird beetle Hip-podamia convergens (Coleoptera Coccinellidae)Can Entomo 113 21-33

Hagen K S 1962 Biology and ecology of predaceousCoccinellidae Annu Rev Entomol 7 289-326

1974 The significance of predaceous Coccinellidae inbiological and integrated control of insects Entomo-phaga 7 25-44

1982 Notes on the convergent lady beetle Univ CalifDiv Agric Nat Resour Leaf 2502

Hagen K S and R R Sluss 1966 Quantity ofaphids required for reproduction by Hippodamia sppin the laboratory pp 47-59 In I Hodek [ed] Ecol-ogy of aphidophagous insects Junk The Hague

Hatch M H and C Tanasse 1948 The liberationof Hippodamia convergens in the Yakima Valley ofWashington 1943 to 1946 J Econ EntomoI 41993

Hodek I 1973 Biology of Coccinellidae Junk TheHague

Hukusima S and T Ohwaki 1972 Further noteson feeding biology of Hannonia axyridis Pallas (Co-leoptera Coccinellidae) Res Bull Fac Agric GifuUniv 33 75--82

694 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

c=J No Beetles - With Beetles A

0250 c==J No Beetles Stored B

- Flown ~ Unmarked200

150

100

50

0Before 15 Days 3 Days

500

-lt1l3 100

300

400

200

bulllt1lC()s1sclaquoWC)+-Ix

B

c

A

80

250200150100

50

- 0~ 50JQ 40a~ 30s~ 20gr 10+I 0

120

Fig 5 Mean + SEM melon aphids per leaf beforeand 3 d after release outdoors of 34-42 adult lady beetlesper pot each pot containing 4 chrysanthemum plants fortrials conducted in May (A) July (B) and September (C)1993 Aphid means in each trial were not significantlydifferent (P gt 001) before release (A) t = 139 n = 1272 (B) t = 031 n = 48 72 (C) t = 008 n = 64 96Means were significantly different (P lt 00001) after re-lease (A) t = 468 n = 20 72 (B) t = 1186 n = 4872 (C) t = 574 n = 6496 Mean plusmn SEM leaves perplant (L) and plant height (H) in centimeters for all treat-ments pooled were (A) L = 154 (04) H = 287 (O(l)(B) L = 158 (04) H = 301 (08) (C) L = 171 (06)H = 297 (11)

40

oBefore Release After Release

Fig 6 Mean + SEM melon aphids per leaf beforeand 15 and 3 d after release outdoors of 40 lown storedor unmarked adult convergent lady beetles per pot eachpot containing 4 chrysanthemum plants for trials con-ducted in July (A) and September (B) 1994 Aphid meansin each trial were not significantly different (P gt 005)before release (A) F = 00 df = 2 213 (B) F = 002df = 3 187 There were significant differences (P lt00001) in aphid densities after release of convergent ladybeetles ANOVA statistics are (A) 15 d F = 195 df =2 69 3 d F = 412 df = 2 141 (B) 15 d F = 146df = 3 92 3 d F = 754 df = 3 116 Mean plusmn SEMleaves per plant (L) and plant height (H) in centimetersfor all treatments pooled were (A) L = 111 (04) H =176 (03) (B) L = 178 (04) H = 252 (04)

Table 1 Melon aphid control provided 15 and 3 dafter a single releasc of 34-42 convergent lady beetlcHper chrysanthemum pot

stored unmarked) (Fig 6B) However at 15 d af-ter release during July 1994 flown beetles provid-ed more control in comparison with stored beetles(Fig 6A) because flown beetles dispersed lessquickly (Fig 3A) so that more individuals werepresent to feed

Discussion

We found that preconditioning aggregation-col-lected H convergens or releasing reared beetles insome instances delayed dispersal in comparisonwith cold-stored beetles from aggregations Ourfindings differ from those of Starks et al (1975)who found no differences in dispersal after release

Mean SE controlMonth Year Treatmenti

15 d 3d

May 1993 All NS 345 (201)July 1993 All NS 784 43)Sept 1993 All NS 844 (59)July 1994 Flown 544 (111) 520 (64)Sept 1994 Flown 289 (147) 686 (63)July 1994 Stored 67 (160) 249 (129)Sept 1994 Stored 353 (116) 636 (58)Sept 1994 Unmarked 539 ( 70) 697 (69)

NS not sampled Calculated according to the Abbott (1925) methodb All all treatments (Hawn stored and reared) were released

on the same plants during 1993 so aphid consumption cannot beseparated by treatment and results from all treatments are com-bined

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 695

40

bull

Fi~ 7 Daily aphid consumption per adult convergentlady wtle (AB aphids per beetle mean t SEM) versusinitial density of melon aphids per leaf (AL aphids perleaf mean t SEM) on potted chrysantl1emums Theequation for the line is AB = 427 (SEM = 123) + 025(SEM = 005) AL (r2 = 082 F = 272 df = 1 6 P =0(02)

density-dependent functional response in preda-tion by H convergens has been demonstrated in amodel by Gutierrez et aI (1981) and has been re-ported for other aphid-feeding lady beetles in-cluding Cocci nella septempunctata L and Har-nwnia axyridis (Pallas) (Hukusima and Ohwaki1972 Hodek 1973)

Hippodamia convergens consumed a mean of 91(SEM = 94) aphids per day which agrees withHodeks (1973) conclusion that the daily feedingrate of adult aphidophagous Coccinellidae usuallyamounts to =100 aphids Clausen (1916) reared IIconvergens in the laboratory on hop aphids Pho-radon humuli (Schrank) which at maturity areabout the same size (1-2 mm body length) as mel-on aphids (Palmer 1952) Clausen found that adultH convergens over their life time consumed =66(SEM = 49 range 17-203) hop aphids each dayClausen used nearly full-grown aphids whereasours were of varying age and included smalleraphids which may account for the 13 lower num-ber of aphids eaten in Clausens (1916) study incomparison with our data Hagen and Sluss (1966)studied H convergens consumption of anothersmall species the spotted alfalfa aphid Therioa-phis maculata (Buckton) Preovipositional adultbeetles consumed 112 (SEM = 21) spotted alfalfaaphids per day and each beetle ate an average of91 aphids per day over their lifetime

Differences in aphid control or dispersal amongtreatments might be more dramatic with differentprerelease treatment of beetles Ignoffo et al(1977) found that clipping the wings of II conver-gens and other entomophages before release re-duced dispersal and increased predation Beetlesin our shldy were active during their 7-10 d out-doors in a screen tent before release (the flowntreatment) beetles moved with daily changes inshade and when disturbed during watering How-ever beetles spent little time actually flying in thetent The main difference between our treatmentswas that flown beetles were exposed to ambientlight wind temperature and humidity for 7-10 dbefore release in comparison with stored beetleswhich were kept at 4-1OdegC and constant dark untilthe day of release

With the exception of 1 lady beetle egg clusterin May 1993 (when reared beetles were released)no reproduction by H convergens was observedGiven rapid dispersal and lack of reproduction anycontrol from a single release would likely be tem-porary Releases are useful for inundative controlbut apparently are not effective for inoculative bi-ological control

Further research is needed to investigate the re-lationships among aphid density beetle releaserates and frequency dispersal and control Ourshldy plants harbored relatively high aphid densi-ties (Figs 5 and 6) and dispersal and control mightdiffer at the lower aphid densities more character-istic of a commercial nursery Hagen (1974) re-ports that the tendency of H convergens to feed

500100 200 300 400AphidsLeaf

oo

in sorghum of nondiapausing native H convergensin comparison with dispersal of aggregation-col-lected beetles Although preconditioning in ourstudy slowed dispersal treatments appear to havelittle practical benefit given that release of bothtreated and untreated beetles controlled aphids H(C)1lVlrglnsproduces few or no eggs unless fedaphids so rearing is expensive Preconditioning ina flight tent requires watering beetles and recol-lecting them before release Some beetles die dur-ing preconditioning reducing the number avail-able at the time of release Because H convergensis relatively inexpensive when purchased in bulk(=$40 buys 70000-80000 beetles) it is less ex-pensive to release greater numbers of stored bee-tles or to introduce them more often to compen-sate for their more rapid dispersal in comparisonwith preconditioning or rearing beetles

Although Starks et al (1975) found no differ-ences in dispersal among treatments native beetlesprovided better aphid control than aggregation-collected beetles When Starks et aI (1975) re-leased 2 beetles per plant in the greenhouse oncaged sorghum infested with 100-400 greenbugsSchizaphis graminum (Rondani) per plant theyobserved =95 and 67 aphid reduction respec-tively 5 d after release of native and aggregation-collected beetles These differences were appar-ently because beetles from aggregations oftenrested at the top of cages without feeding onaphids

Hippodamia convergens in our study consumed=25-170 melon aphids per day increasing con-sumption as aphid density increased (Fig 7) This

160 bull 19930 1994

Q) 120

fQ)Q)al-()C 80Ea4

696 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

aggregate and disperse varies with seasonal differ-ences in physiological condition of the beetlesHost plant aphid species release environmentand environmental conditions also may affectaphid control from beetle release

We are not aware of any studies indicating thatconvergent lady beetle releases are effective incontrolling aphids on a large scale Our preliminaryresearch on landscape roses indicates that releaseof relatively large numbers of beetles or multiplereleases may be needed to obtain control on largerplants Little is known about overwintering H con-vergens and the potential ecological effects oflarge-scale harvesting of aggregating lady beetleshave not been investigated Even if releases of ag-gregation-collected convergent lady beetles couldcontrol aphids on a large scale potential ecologicalor nontarget effects might make large-scale releas-es undesirable

There are quarantine considerations regardingthe shipment of field-collected insects Aggregat-ing beetles may harbor microorganisms and bulkcollections may be contaminated with other inver-tebrates The parasite Dinocampus (=Perilitus)coccinellae (Shrank) (Hymenoptera Braconidae)emerged from a few of the commercial H conver-gens we received We did not dissect hosts to checkparasitism and did not hold beetles and feed themaphids to induce D coccinellae emergence so wedo not know the extent of their parasitization byD coccinellae which attacks gt40 coccinellid spe-cies (Obrycki 1989) Ruzicka and Hagen (1985)found that 10 of the H convergens collectedfrom aggregations in California are parasitized byD coccinellae and parasitized beetles have less ofa tendency to disperse in comparison with unpar-asitized H convergens

Augmentative release of commercially availableconvergent lady beetles can provide inundativecontrol of relatively high aphid densities on sIllallpotted plants in a limited area Convergent ladybeetles are organically acceptable readily avail-able hardy easily handled and store well Pesti-cide exposure hazards and worker reentry restric-tions after pesticide application may make ladybeetle releases an attractive option for aphid con-trol in nurseries and greenhouses However ladybeetle releases must be integrated with other man-agement practices and pesticide applications forother pests may need to be modified for releasesto be effective

Acknowledgments

Michael P Parrella Norman E Gary Jill RentnerChuck LeGault and Bob Rousseau (Department of En-tomology University of California Davis) and Robin Ro-setta Pat Thompson and Ellen Zagory (Davis Arbore-tum University of California Davis) assisted usKenneth S Hagen and Kent M Daane (Laboratory ofBiological Control University of California Berkeley)Lester E Ehler (Department of Entomology Universityof California Davis) and Michael J Raupp (Department

of Entomology University of Maryland) critiqued draftsof the manuscript Unique Insect Control CitrusHeights CA provided the lady beetles The Universityof California Elvenia J Slosson Fund for OrnamentalHorticulture provided financial assistance

References Cited

Abbott W S 1925 A method of computing the ef-fectiveness of an insecticide J Econ Entomol 18265-267

Clausen C P 1916 Life-history and feeding recordsof a series of California Coccinellidae Univ CalifPub Entomol 1(6) 251-299

Cooke W C 1963 Ecology of the pea aphid in theblue mountain area of eastern Oregon and Washing-ton US Dep Agric Tech Bull 1287

Davidson W M 1919 The convergent ladybirdbeetle (Hippodamia convergens Guerin) and the bar-ley-com aphis (Aphis lIuzidis Fitch) Mon Bull CalifState Comm Homc 8 23-26

1924 Observations and experiments on the dispersionof the convergent lady-beetle (Hippodamia conver-gens Guerin-MenevilIe) in California Trans Am En-tomol Soc 50 163--175

Davis J R and R L Kirkland 1982 Physiologicaland environmental factors related to the dispersalflight of the convergent lady beetle Hippodamia con-vergens (Guerin-MenevilIe) J Kans Entomo Soc55 187-196

Eddy C O 1939 An attempt to colonize Hippodamiaconvergens Proc Int Soc Sugar Cane Technol 6385--386 (abstracted in Rev App Entomo 28 245[1940])

Fenton F A and R G Dahms 1951 Attempts atcontrolling the greenbug by the importation and re-lease of lady beetles in Oklahoma Proc Okla AcadSci 32 49-51

Flint M L S H Dreistadt J Rentner and M PParrella 1995 Lady beetle release controls aphidson potted plants Calif Agric 49(2) 5--8

Gutierrez A P J U Baumgaertner and K S Ha-gen 1981 A conceptual model for growth devel-opment and reproduction in the ladybird beetle Hip-podamia convergens (Coleoptera Coccinellidae)Can Entomo 113 21-33

Hagen K S 1962 Biology and ecology of predaceousCoccinellidae Annu Rev Entomol 7 289-326

1974 The significance of predaceous Coccinellidae inbiological and integrated control of insects Entomo-phaga 7 25-44

1982 Notes on the convergent lady beetle Univ CalifDiv Agric Nat Resour Leaf 2502

Hagen K S and R R Sluss 1966 Quantity ofaphids required for reproduction by Hippodamia sppin the laboratory pp 47-59 In I Hodek [ed] Ecol-ogy of aphidophagous insects Junk The Hague

Hatch M H and C Tanasse 1948 The liberationof Hippodamia convergens in the Yakima Valley ofWashington 1943 to 1946 J Econ EntomoI 41993

Hodek I 1973 Biology of Coccinellidae Junk TheHague

Hukusima S and T Ohwaki 1972 Further noteson feeding biology of Hannonia axyridis Pallas (Co-leoptera Coccinellidae) Res Bull Fac Agric GifuUniv 33 75--82

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 695

40

bull

Fi~ 7 Daily aphid consumption per adult convergentlady wtle (AB aphids per beetle mean t SEM) versusinitial density of melon aphids per leaf (AL aphids perleaf mean t SEM) on potted chrysantl1emums Theequation for the line is AB = 427 (SEM = 123) + 025(SEM = 005) AL (r2 = 082 F = 272 df = 1 6 P =0(02)

density-dependent functional response in preda-tion by H convergens has been demonstrated in amodel by Gutierrez et aI (1981) and has been re-ported for other aphid-feeding lady beetles in-cluding Cocci nella septempunctata L and Har-nwnia axyridis (Pallas) (Hukusima and Ohwaki1972 Hodek 1973)

Hippodamia convergens consumed a mean of 91(SEM = 94) aphids per day which agrees withHodeks (1973) conclusion that the daily feedingrate of adult aphidophagous Coccinellidae usuallyamounts to =100 aphids Clausen (1916) reared IIconvergens in the laboratory on hop aphids Pho-radon humuli (Schrank) which at maturity areabout the same size (1-2 mm body length) as mel-on aphids (Palmer 1952) Clausen found that adultH convergens over their life time consumed =66(SEM = 49 range 17-203) hop aphids each dayClausen used nearly full-grown aphids whereasours were of varying age and included smalleraphids which may account for the 13 lower num-ber of aphids eaten in Clausens (1916) study incomparison with our data Hagen and Sluss (1966)studied H convergens consumption of anothersmall species the spotted alfalfa aphid Therioa-phis maculata (Buckton) Preovipositional adultbeetles consumed 112 (SEM = 21) spotted alfalfaaphids per day and each beetle ate an average of91 aphids per day over their lifetime

Differences in aphid control or dispersal amongtreatments might be more dramatic with differentprerelease treatment of beetles Ignoffo et al(1977) found that clipping the wings of II conver-gens and other entomophages before release re-duced dispersal and increased predation Beetlesin our shldy were active during their 7-10 d out-doors in a screen tent before release (the flowntreatment) beetles moved with daily changes inshade and when disturbed during watering How-ever beetles spent little time actually flying in thetent The main difference between our treatmentswas that flown beetles were exposed to ambientlight wind temperature and humidity for 7-10 dbefore release in comparison with stored beetleswhich were kept at 4-1OdegC and constant dark untilthe day of release

With the exception of 1 lady beetle egg clusterin May 1993 (when reared beetles were released)no reproduction by H convergens was observedGiven rapid dispersal and lack of reproduction anycontrol from a single release would likely be tem-porary Releases are useful for inundative controlbut apparently are not effective for inoculative bi-ological control

Further research is needed to investigate the re-lationships among aphid density beetle releaserates and frequency dispersal and control Ourshldy plants harbored relatively high aphid densi-ties (Figs 5 and 6) and dispersal and control mightdiffer at the lower aphid densities more character-istic of a commercial nursery Hagen (1974) re-ports that the tendency of H convergens to feed

500100 200 300 400AphidsLeaf

oo

in sorghum of nondiapausing native H convergensin comparison with dispersal of aggregation-col-lected beetles Although preconditioning in ourstudy slowed dispersal treatments appear to havelittle practical benefit given that release of bothtreated and untreated beetles controlled aphids H(C)1lVlrglnsproduces few or no eggs unless fedaphids so rearing is expensive Preconditioning ina flight tent requires watering beetles and recol-lecting them before release Some beetles die dur-ing preconditioning reducing the number avail-able at the time of release Because H convergensis relatively inexpensive when purchased in bulk(=$40 buys 70000-80000 beetles) it is less ex-pensive to release greater numbers of stored bee-tles or to introduce them more often to compen-sate for their more rapid dispersal in comparisonwith preconditioning or rearing beetles

Although Starks et al (1975) found no differ-ences in dispersal among treatments native beetlesprovided better aphid control than aggregation-collected beetles When Starks et aI (1975) re-leased 2 beetles per plant in the greenhouse oncaged sorghum infested with 100-400 greenbugsSchizaphis graminum (Rondani) per plant theyobserved =95 and 67 aphid reduction respec-tively 5 d after release of native and aggregation-collected beetles These differences were appar-ently because beetles from aggregations oftenrested at the top of cages without feeding onaphids

Hippodamia convergens in our study consumed=25-170 melon aphids per day increasing con-sumption as aphid density increased (Fig 7) This

160 bull 19930 1994

Q) 120

fQ)Q)al-()C 80Ea4

696 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

aggregate and disperse varies with seasonal differ-ences in physiological condition of the beetlesHost plant aphid species release environmentand environmental conditions also may affectaphid control from beetle release

We are not aware of any studies indicating thatconvergent lady beetle releases are effective incontrolling aphids on a large scale Our preliminaryresearch on landscape roses indicates that releaseof relatively large numbers of beetles or multiplereleases may be needed to obtain control on largerplants Little is known about overwintering H con-vergens and the potential ecological effects oflarge-scale harvesting of aggregating lady beetleshave not been investigated Even if releases of ag-gregation-collected convergent lady beetles couldcontrol aphids on a large scale potential ecologicalor nontarget effects might make large-scale releas-es undesirable

There are quarantine considerations regardingthe shipment of field-collected insects Aggregat-ing beetles may harbor microorganisms and bulkcollections may be contaminated with other inver-tebrates The parasite Dinocampus (=Perilitus)coccinellae (Shrank) (Hymenoptera Braconidae)emerged from a few of the commercial H conver-gens we received We did not dissect hosts to checkparasitism and did not hold beetles and feed themaphids to induce D coccinellae emergence so wedo not know the extent of their parasitization byD coccinellae which attacks gt40 coccinellid spe-cies (Obrycki 1989) Ruzicka and Hagen (1985)found that 10 of the H convergens collectedfrom aggregations in California are parasitized byD coccinellae and parasitized beetles have less ofa tendency to disperse in comparison with unpar-asitized H convergens

Augmentative release of commercially availableconvergent lady beetles can provide inundativecontrol of relatively high aphid densities on sIllallpotted plants in a limited area Convergent ladybeetles are organically acceptable readily avail-able hardy easily handled and store well Pesti-cide exposure hazards and worker reentry restric-tions after pesticide application may make ladybeetle releases an attractive option for aphid con-trol in nurseries and greenhouses However ladybeetle releases must be integrated with other man-agement practices and pesticide applications forother pests may need to be modified for releasesto be effective

Acknowledgments

Michael P Parrella Norman E Gary Jill RentnerChuck LeGault and Bob Rousseau (Department of En-tomology University of California Davis) and Robin Ro-setta Pat Thompson and Ellen Zagory (Davis Arbore-tum University of California Davis) assisted usKenneth S Hagen and Kent M Daane (Laboratory ofBiological Control University of California Berkeley)Lester E Ehler (Department of Entomology Universityof California Davis) and Michael J Raupp (Department

of Entomology University of Maryland) critiqued draftsof the manuscript Unique Insect Control CitrusHeights CA provided the lady beetles The Universityof California Elvenia J Slosson Fund for OrnamentalHorticulture provided financial assistance

References Cited

Abbott W S 1925 A method of computing the ef-fectiveness of an insecticide J Econ Entomol 18265-267

Clausen C P 1916 Life-history and feeding recordsof a series of California Coccinellidae Univ CalifPub Entomol 1(6) 251-299

Cooke W C 1963 Ecology of the pea aphid in theblue mountain area of eastern Oregon and Washing-ton US Dep Agric Tech Bull 1287

Davidson W M 1919 The convergent ladybirdbeetle (Hippodamia convergens Guerin) and the bar-ley-com aphis (Aphis lIuzidis Fitch) Mon Bull CalifState Comm Homc 8 23-26

1924 Observations and experiments on the dispersionof the convergent lady-beetle (Hippodamia conver-gens Guerin-MenevilIe) in California Trans Am En-tomol Soc 50 163--175

Davis J R and R L Kirkland 1982 Physiologicaland environmental factors related to the dispersalflight of the convergent lady beetle Hippodamia con-vergens (Guerin-MenevilIe) J Kans Entomo Soc55 187-196

Eddy C O 1939 An attempt to colonize Hippodamiaconvergens Proc Int Soc Sugar Cane Technol 6385--386 (abstracted in Rev App Entomo 28 245[1940])

Fenton F A and R G Dahms 1951 Attempts atcontrolling the greenbug by the importation and re-lease of lady beetles in Oklahoma Proc Okla AcadSci 32 49-51

Flint M L S H Dreistadt J Rentner and M PParrella 1995 Lady beetle release controls aphidson potted plants Calif Agric 49(2) 5--8

Gutierrez A P J U Baumgaertner and K S Ha-gen 1981 A conceptual model for growth devel-opment and reproduction in the ladybird beetle Hip-podamia convergens (Coleoptera Coccinellidae)Can Entomo 113 21-33

Hagen K S 1962 Biology and ecology of predaceousCoccinellidae Annu Rev Entomol 7 289-326

1974 The significance of predaceous Coccinellidae inbiological and integrated control of insects Entomo-phaga 7 25-44

1982 Notes on the convergent lady beetle Univ CalifDiv Agric Nat Resour Leaf 2502

Hagen K S and R R Sluss 1966 Quantity ofaphids required for reproduction by Hippodamia sppin the laboratory pp 47-59 In I Hodek [ed] Ecol-ogy of aphidophagous insects Junk The Hague

Hatch M H and C Tanasse 1948 The liberationof Hippodamia convergens in the Yakima Valley ofWashington 1943 to 1946 J Econ EntomoI 41993

Hodek I 1973 Biology of Coccinellidae Junk TheHague

Hukusima S and T Ohwaki 1972 Further noteson feeding biology of Hannonia axyridis Pallas (Co-leoptera Coccinellidae) Res Bull Fac Agric GifuUniv 33 75--82

696 ENVIRONMENTAL ENTOMOLOGY Vol 25 no 3

aggregate and disperse varies with seasonal differ-ences in physiological condition of the beetlesHost plant aphid species release environmentand environmental conditions also may affectaphid control from beetle release

We are not aware of any studies indicating thatconvergent lady beetle releases are effective incontrolling aphids on a large scale Our preliminaryresearch on landscape roses indicates that releaseof relatively large numbers of beetles or multiplereleases may be needed to obtain control on largerplants Little is known about overwintering H con-vergens and the potential ecological effects oflarge-scale harvesting of aggregating lady beetleshave not been investigated Even if releases of ag-gregation-collected convergent lady beetles couldcontrol aphids on a large scale potential ecologicalor nontarget effects might make large-scale releas-es undesirable

There are quarantine considerations regardingthe shipment of field-collected insects Aggregat-ing beetles may harbor microorganisms and bulkcollections may be contaminated with other inver-tebrates The parasite Dinocampus (=Perilitus)coccinellae (Shrank) (Hymenoptera Braconidae)emerged from a few of the commercial H conver-gens we received We did not dissect hosts to checkparasitism and did not hold beetles and feed themaphids to induce D coccinellae emergence so wedo not know the extent of their parasitization byD coccinellae which attacks gt40 coccinellid spe-cies (Obrycki 1989) Ruzicka and Hagen (1985)found that 10 of the H convergens collectedfrom aggregations in California are parasitized byD coccinellae and parasitized beetles have less ofa tendency to disperse in comparison with unpar-asitized H convergens

Augmentative release of commercially availableconvergent lady beetles can provide inundativecontrol of relatively high aphid densities on sIllallpotted plants in a limited area Convergent ladybeetles are organically acceptable readily avail-able hardy easily handled and store well Pesti-cide exposure hazards and worker reentry restric-tions after pesticide application may make ladybeetle releases an attractive option for aphid con-trol in nurseries and greenhouses However ladybeetle releases must be integrated with other man-agement practices and pesticide applications forother pests may need to be modified for releasesto be effective

Acknowledgments

Michael P Parrella Norman E Gary Jill RentnerChuck LeGault and Bob Rousseau (Department of En-tomology University of California Davis) and Robin Ro-setta Pat Thompson and Ellen Zagory (Davis Arbore-tum University of California Davis) assisted usKenneth S Hagen and Kent M Daane (Laboratory ofBiological Control University of California Berkeley)Lester E Ehler (Department of Entomology Universityof California Davis) and Michael J Raupp (Department

of Entomology University of Maryland) critiqued draftsof the manuscript Unique Insect Control CitrusHeights CA provided the lady beetles The Universityof California Elvenia J Slosson Fund for OrnamentalHorticulture provided financial assistance

References Cited

Abbott W S 1925 A method of computing the ef-fectiveness of an insecticide J Econ Entomol 18265-267

Clausen C P 1916 Life-history and feeding recordsof a series of California Coccinellidae Univ CalifPub Entomol 1(6) 251-299

Cooke W C 1963 Ecology of the pea aphid in theblue mountain area of eastern Oregon and Washing-ton US Dep Agric Tech Bull 1287

Davidson W M 1919 The convergent ladybirdbeetle (Hippodamia convergens Guerin) and the bar-ley-com aphis (Aphis lIuzidis Fitch) Mon Bull CalifState Comm Homc 8 23-26

1924 Observations and experiments on the dispersionof the convergent lady-beetle (Hippodamia conver-gens Guerin-MenevilIe) in California Trans Am En-tomol Soc 50 163--175

Davis J R and R L Kirkland 1982 Physiologicaland environmental factors related to the dispersalflight of the convergent lady beetle Hippodamia con-vergens (Guerin-MenevilIe) J Kans Entomo Soc55 187-196

Eddy C O 1939 An attempt to colonize Hippodamiaconvergens Proc Int Soc Sugar Cane Technol 6385--386 (abstracted in Rev App Entomo 28 245[1940])

Fenton F A and R G Dahms 1951 Attempts atcontrolling the greenbug by the importation and re-lease of lady beetles in Oklahoma Proc Okla AcadSci 32 49-51

Flint M L S H Dreistadt J Rentner and M PParrella 1995 Lady beetle release controls aphidson potted plants Calif Agric 49(2) 5--8

Gutierrez A P J U Baumgaertner and K S Ha-gen 1981 A conceptual model for growth devel-opment and reproduction in the ladybird beetle Hip-podamia convergens (Coleoptera Coccinellidae)Can Entomo 113 21-33

Hagen K S 1962 Biology and ecology of predaceousCoccinellidae Annu Rev Entomol 7 289-326

1974 The significance of predaceous Coccinellidae inbiological and integrated control of insects Entomo-phaga 7 25-44

1982 Notes on the convergent lady beetle Univ CalifDiv Agric Nat Resour Leaf 2502

Hagen K S and R R Sluss 1966 Quantity ofaphids required for reproduction by Hippodamia sppin the laboratory pp 47-59 In I Hodek [ed] Ecol-ogy of aphidophagous insects Junk The Hague

Hatch M H and C Tanasse 1948 The liberationof Hippodamia convergens in the Yakima Valley ofWashington 1943 to 1946 J Econ EntomoI 41993

Hodek I 1973 Biology of Coccinellidae Junk TheHague

Hukusima S and T Ohwaki 1972 Further noteson feeding biology of Hannonia axyridis Pallas (Co-leoptera Coccinellidae) Res Bull Fac Agric GifuUniv 33 75--82

June 1996 DREISTADT AND FLINT LADY BEETLE RELEASE 697

I~10ffo C M C Garcia W A Dickerson G TSchmidt lind K D Biever 1977 Imprisonmentof entomophages to increase effectiveness evaluationof a concept J Econ Entomol 70 292-294

Moore W S and C S Koehler 1981 Aphids inthe home garden and landscape Univ Calif DivAgric Nat Resour LeaR 21032

Obrycki J J 1989 Parasitization of native and exoticcoccinellids by Dinocampus coccinellae (Schrank)(Hymenoptera Braconidae) J Kans Entomol Soc62 211-218

Packard C M and R E CantpbeU 1926 The peaaphid as an alfalfa pest in California J Econ Ento-mol 19 752-761

Palmer M A 1952 Aphids of the Rocky Mountainregion Thomas Say Foundation 5 Entomological So-ciety of America Lrnlham MD

RlInkin M A and S Rankin 1980 Some factorsaffecting presumed migratory flight activity of theconvergent lady beetle Hippodamia convergens(Coccinellidae Coleoptera) BioI Bull 158 356-369

Raupp M J M R Hardin S M Braxton and BB Bull 1994 Augmentative releases for aphid con-trol on landscape plants J Arboric 20 241-249

Ruzicka Z and K S Hagen 1985 Impact of par-asitism on migratory flight performance in females ofHippodamia convergens (Coleoptera) Acta EntomolBohemoslov 82 401-406

SAS Institute 1988 SASSTAT users guide release603 SAS Institute Cary NG

Starks K J E A Wood R L Burton and H WSomsen 1975 Behavior of convergent lady beetlesin relation to greenbug control in sorghum US DepAgric Agric Res ServoS-53

Vehrs SLC G P Walker and M P PllrreUa 1992Comparison of population growth rate and within-plantdistribution between Aphis gossypii and Myzus persi-cae (Homoptera Aphididae) reared on potted chrysan-themums J Econ Entomol 85 799-807

Received for publication 8 May 1995 accepted 30 Jan-uary 1996