S O U T IIW E S T E R N E N T O M O L O G IS T N IA R .2003

M ttIIN ‐F IE L D D IST R IB U T IO N O F T H R E E H O M O PT E R A N SP E C ttS IN T E X A S SU G A R C A N E

R. L. Meagher,Jr.r andJ. C. kgaspi 2

TexasA&M University Agricultural Researchand ExtensionCenter 2415EastHighway 83 Weslaco,Texas 78596

ABSTRACT

Sugarcanefields composed ofeither'CP 70-321'or'NCo 310'weresampled during 1993 and 1994 for three speciesofhomopteran insects. The West Indian canefly, Saccharosydne saccharivora(tlestwood), wasthe mostabundant speiies collectedwith densitiesreaching over 40 per shoot. Populationdensities varied sigrificantly throughoutthe seasonin 1993but were similar in 1994. No differencein densitieswere found between sugarcane cultivars. Both the sugarcanedelphacid, Perhinsiella saccharicidaKirkaldyand the leafhopperDraeculacephala portola Ball, were found in low numbers(< 1.0per shoot). P. saccharicidareached its highest levelslater in the season,and 'CP 70-321'shoots harbored more individuals than 'NCo 310' shoots.Although D. portola dertsitieswere low, differe,ncesamong fields andbetwecn cultivars were found. Aggregation,as measuredusing Taylor a and b coefficients, was shown to be greaterfor ,S.saccharivora than the other species.

INTRODUCTION

Sugarcane(interspecific hybrids of Saccharum)has been commercially grown in a tlree- county areaof southemTexas since 1972. Stemboringpyralids, Mexican rice borer, Eoreuma loftini (Dyar), andsugarcane borer, Diatraea saccharalis(F.), havebeen the most seriousinsect pests of the in{ustry (Meagher et al. 1994); however, other insects are active in the Texas sugarcaneagroecos)Nstem. The homopteranfauna was described in two earlierreports (Meagh€r et al. 1991,Meagher et al. 1993),although only basic surveyinformation was noted and samplingwas conductedusing a large suction device that did not estimateinsects per shoot. Three Auchenorrhyncha homopteran species, Saccharosydnesaccharivora (Westwood) (Delphacidae),Perhinsiella saccharicidaKirkaldy (Delphacidae), and Draeculacephala portola Ball (Cicadellidae),were relatively abundantin theseearlier surveys. l C u r ent ad dress:U S D A ‐A R S C M A V E , 1700 S W 2 3rd D r.,G おn es宙1le,F L 3260 8 2 ― c u rren t addressi U S D A A R S C M A V E , C enter for B lo logical C ontro l, F loi da A & M U n iv ersllんT allah assee,F L 32 307 The West Indian canefly,,Saccharosydne saccharivora,hasbeen associatedwith sugarcane in the Caribbean and in North, South, and Cenffal America for centuries (Metcalfe 1969). This delphacid has historically caused va.rying degrees of economic damage (Charpentier 1970), and has been documented in all continental sugarcane-producing states (Charpentier I 970, Hall I 988, Meagher et al. 1993). The sugarcane delphacid, P. saccharicida, hrst noted in the continental United States in Florida in 1982 (Sosa 1985), was documented in Texas in 1989 (Meagher et al. 1991) and was recently discovered in Louisiana in 1994 (White et al. 1995). This insect can cause economic damage (Wilson 1987) and is a vector of Fijivirus sp., the causal agent of Fiji disease of sugarcane (Francki and Grivell 1972,Egan et al. 1989). D. portola is commonly collected in sugarcane (Hall 1988, Meagher et al. 1993) and is sporadically captured in other grass crops (Wilson et al. 1973, Hawkins et al. I 979). It was reported to be a vector of chlorotic streak in sugarcane (Abbott and Ingram 1942) until further investigation proved otherwise (Abbou et al. 1961). An understanding of the spatial pattems of an insect can aid in developing sampling plans (Southwood 1978). Spatial pattems are affected by intrinsic factors such as oviposition pattem and immature dispersal, extrinsic factors such as host quality and environmental toxicants, and estimation procedures such as sampling plan paf,ameters,sample unit size, number of samples, and sampler bias (Wilson 1994). Spatial pattems can depend on population density, with low densities tending to be indistinguishable from uniform or random, and high densities of most insects tending to be aggregated. The efficiency of a resulting sampling plan must balance cost considerations against the quality of the information provided. The objective of the present research was to describe the within-field distributions of S. saccharivora, P. saccharicida, and D. portola adults and nymphs in two sugarcane cultivars and calculate sample sizes based on a defined level of reliabilitv.

MATERIALS AND METHODS

'CP Fields and Sampling. Commercial sugarc:me fields of either 70-321' or'NCo 310' in different Lower Rio Grande Valley locations were sampled in 1993 ard 1994. Field size ranged from 6.3 to 15.7 ha. Sampling was initiated during February or March and terminated in late August due to sugarcane harvesting. Every effort was taken to maintain a biweekly timetable; however, weather conditions or irrigation schedules occasionally delayed sampling in individual fields. Each field was sampled a total of 8 to 10 times per year. Sampling was conducted duringmid- to late morningbyrandomlyselecting 30 shoots (randomizedbyselecting sugarcane row and number of paces within row) per field. Starting at the base of the shoot, leaves were gently pulled back and the number of adults and nymphs present was recorded. Statistical Analyiis. Means per shoot for each species were calculated for each weekly sample taken in each field. Numbers were compared among weeks, among fields and between cultivars using analysis of variance @ROC GLM, SAS lnstitute 1996). Counts were square root (y + 0.5) transformed before analysis, but untransformed means are shoyn in taQles and figwes. b Spatial dispersion was investigated using Taylor coefficients, 32 : a * lTaylor'i961, 1984). Taylor coefficients were estimated by frtting ln (S-) : ln a + b ln ( x) (PROC REG, SAS Institute 1996). Estimates for d and D were compared using analyses of variance (PROC GLM, SAS Institute 1996) to examine the effects of cultivar. Wilson and Room (1982) presented the following equation for estimating sample size where the goal was to estimate population density with a defined level of reliability: (td2/Dr)2 o-(b-'), where 1.,72is the standard normal variate for a two-tailed confidence interval, D" is a proportion defined as the ratio of half the desired CI to the mean (D" : lCV2l / x for enumerative sampling), and a and D are Taylor coefficients. This equation permits one to estimate required sample size (n) over a range of densities for any species and sample unit whose Taylor coefficientsare known. Requiredsample size was calculated using ta/2-- 1.645and D" :0.2.

RESLILTSAND DISCUSSION

Higher numbersof S.saccharivora were collectedin 1993 than I 994,with seasonalmeans of23.7 and3.0individualspershoot,respectively.Populationsvariedsignificantlythroughout the seasonin I 993 (F -- 3.2; df : 23,42; P : 0.0005),with peakpopulations occurring in May (Fig. 1a). Meagheret al. (1993)also found more S.sacchaivora during May-Junethan March- April. Migrating adultsinto youngor ratoon sugzrcanefields can boostpopulations !o over 30 per shootduring the initial phasesof colonization(Metcalfe 1969), and this wasexemplified in onefieldwithameanof44.T+7.8(SE)Westhrdiancaneflypershoot.Theselargepopulations in a 'NCo 310' field nearElsa contributed to significantamong field variation(F: 6.9; df :3, 42; P = 0.QQ07).However, S. saccharivora were found in similar numbersbetween cultivars (F : l.8; df = 1,42; P:0.1898). kr 1994,only amongfield variationwas sigrificant (F -- 2.9; df : 4,3l; P = 0.0403)(Fig. lb). S. saccharivoraadults and nymphs were the leastmobile and wereeasy to locatebecause they were usually present on theunderside ofleafblades. Perbinsiellasaccharicidapopulations were low inboth years,not exceeding1.0 pershoot. Low numbers of P. saccharicida wqe found in our previous study; however,sampling was conductedusinga largesuction sampling method thatprobablywasn't appropriate forthis insect (Meagfueretal.1993). lnl993,populationsvariedsignificantlythroughouttheseason(F:3.0' df:23,42;P:0.OOl),withpeakpopulationsoccur4inginJulyandAugust(Fig.2a).Among field and cultivar differenceswere not significant (P > 0.05). Seasonalabundance studies in Floridashowed that sugarcane delphacid population densities ranged widelyfield-to-field (Sosa 1985)and increased during the summermonths with peakdensities approaching 7.5 per shoot in Octoberand November(Sosa et al. 1986). However,maximum population densities in Louisianasugarcane were found to be 0.3 per shoot(White et al. 1995). We werenot ableto sampleduring fall and winter due to sugarcaneharvesting. Peakpopulation densitiesof this insectin southeastemQueensland, Australia were as high as96 adultsand 335 nymphs per shoot (Allsoppand Bull 1990). br 1994,seasonal, among field, and cultivar variationwere all significant(P < 0.05). Populationdensities were low from Februarythrough April, but were higher from May through August (Fig. 2b). 'CP 70-321' shootscontained more sugarcanedelphacid adults and nymphs 'NCo than 310' shoots(F: 5.6; df : 1, 3l; P:0.0242) (Fig. 2c). Cultivarpreference and populationdevelopment differences have been shown experimentally and in field studies(Chang andOta 1978,Taniguchi et al. 1980,Allsopp andBull 1990). Densitiesfor D. portola werealso low, neverexceeding L0 per shoot. Earlierresearch showedthis speciesto be the most cornmonlycollected (Meagher et al. 1993). However, samplesin that studywere takenfrom sugarcaneand surroundinggrasses using a largesuction sampler, and it is possible that most of the D. portola population was collected from non- sugarcanehosts. Populationdensities in both yearswere not different acrossweeks (P > 0.14) but weredifferent amongfields andbetween cultivars (P < 0.05). kr 1993,densities were higher 'NCo 'CP in 310' than 70-321'shoots (0.2 * 0.01vs.0.12 + 0.01,respectively; F = 14.5;df : 1,42; P: 0.0004)(Figs. 3a and 3b), while in 1994,'CP 70-321'shoots harbored more 'NCo leaftoppersthan 310'shoots (0.37 + g.g2ur.6.27 *0.O2,respectively; F= 4.4;df :1,31; P:0.0444) @igs.4a and4b). D. ponola werethe most active species collected, with nymphs presentin the whorls and adultsin whorls and on leafblades. Tayrora coefficientranged from 0.85 for D. portola to 4.15 for s. saccharivora,and b rangedfrom 0.95 for D. portola to 1.55for ^S.saccharivora (Table l). Taylorcoefficients for P. saccharicidawere in betweenthose values and were similar to coeffrcientscalculated forP. saccharicidanymphs and adults sampled in Australiansugarcane fields (Allsopp andBull 1990). Taylorbcoefficientwassignificantly>1.0forS.saccharivorainbothyearsandP.saccharicida ― ― □ E lsa‐3 10

- ― △ L aF 9古a‐3 10

- M ercedes‐3 10

- ― ◇ S an Ju an ‐3 10

- r E E lsa‐32 1

~ ~ ▲ L aF9占a‐32 1

- ― ● M ercedes-32 1

- ― ◆ San Juan‐32 1

Ｏ ０ ｈ ｈ 】 ｈ あ あ あ ぁ あ ロ 自 宮 日 【 【 【 【 的 的 的 的 的 ● ● ａ ａ ａ ａ 田 電 田 付 田 ヨ ョ ョ ョ ョ Ｆ ョ ヨ 宮 巧 ヨ ヨ ヨ ヨ ヨ 』 Ｌ く く 才 Ｈ 噂 専 噂 専 ， ち 均 ヽ ヽ ， ヽ く 手 く く く ヽ 【宮 宮 宮 百 百 崎 ヽ ト 寺 ∞ 崎 い ６ 〕 ヽ 崎 い ６ 〕 ● 卜 【 【 い い 〕 的 ｏ Ｈ Ｎ 【 崎 や Ｈ Ｈ 【 Ｎ Ｈ 【 ま Ｈ 宮 Ｈ い ｍ Ｈ 【 輸

° b 。 ~ ~ 口 A lam o‐3 10

~ ~ △ E lsa‐3 10 8 - ― や O L aF tt a‐3 10 Ｏ Ｏ 〓 - ◇一 , り M ercedes 3 10 】 ● - 日― ‐ ａ E lsa 32 1 ロ ~ ~ ミ ▲ L aF tt a‐32 1 ０ 日 宮 - - 0 M ercedes‐32 1

- ― ◆ R unn ‐32 1

gg:g:自3333富富富富富gigミ尋暑呂景ゴss W eek

FIG . 1, Seasonal density oR ttcc力α rO串品 夕sα“力αガソθ″ in eight sugarcane n elds,1。w er R lo ` ' G rande V alley,T exas,1993 and 1994 . T he densiけ On 23 M ay 1993 w as 44.7 per shoot. 3 10 ` ' refers to cuttivar N C o 3 10, 32 1 refers to culuvar c P 70-32 1. 0・9° a . 0.80

0.70 貿 ｏ ０６０ ● ， 〓 り Ｌ 蜘 ● ａ 宮 伽 日 ０ 一 々 蜘 伽

00 0 :ミ景s畳患富冒gg言長ξ長昂言言昌景富FSSS 0.40 b. ~ 1994 八 いAlamo NCo 310/ -^- 〕 \ ,r,u ｏ ｏ ａ り Ｌ ● 付 宮 電 ０ 日 を

:富3:畳景富貫g言旨這:言g尋:昌景冒冨s

0潟0 C . ｍ ~い Eヽa l1 19 9 4 ｍ や -^- CP 70'321 Ｏ LaFeria ^/ \ Ｏ 〓 蜘 り Ｌ Ｏ ａ 蜘 宮 口 ０ ０３０ 日 々 ０・２０ ０・１０ ｍ gg言:::33畳景富富g言冒逼::ミ尋昌昌目ぎ W eek

FIG .2.Seasonal density ofFセrttzs,97ra sacc力αァic,】αin sugarcane for 1993 and 1994,low erR lo G rande V alleL Texas, M eans in 1993 are across bodl cuitivars and all eight ields. M eans in 1994 are for four i elds for each Ofthe cultivars N C o 310 and C P 70-321. ・°° a . ~ ー ロ E lsa N C o 3 10 一 一 0.80 ▲ L aF erla や ｏ ~ ~ ｏ O M erccde s ヨ ０ ６０ リ 一 一 出 ◆ S an Ju an ０ 亀 宮 ミ ０ ４０ ● 壇 “

０ ０ 」 ｈ 】 」 浄 あ 静 ぁ ヽ 宮 ョ 目 ロ 【 【 【 【 的 的 的 的 的 ● ● 亀 ａ 虫 ａ 韓 韓 電 再 ョ ョ ョ ョ ョ ョ ヨ 』 ” ， ヽ ， ヨ ヨ ヨ ヨ 白 く く く く Σ 甲 濯 噂 噂 ， ヽ ヽ ヽ 巧 く く く ど く 百 宮 宮 崎 い 崎 い ６ 〕 ヽ 崎 い ６ 〕 ト 寺 【 ∞ ヽ い じ ｏ 出 い 碕 ● 」 寺 一 Ｈ ヽ 田 〓 百 ヽ ” Ｈ 日 田 Ｈ 【 ” Ｈ 【 的

W eek .50 b . C P 70‐321 0。40 _▲_ や ０ 一 一 ● 〇 ヨ ０ ３０ リ ． Ｌ Ｏ

ａ

ロ ミ ０ ２０ ● ． 専 宮

０ １０ ．

ｎ ● 」 ｈ ｈ ｈ あ ぁ ヽ ぁ あ Ｈ 日 宮 ロ 【 【 】 ≡ 的 的 的 ● ● ら ａ ａ ａ 苺 『 電 毎 車 ョ ョ ョ ョ ョ 再 ヨ ョ ヽ 〜 ヨ ヨ ヨ 』 比 く く く く Ｈ 日 専 専 】 巧 ヽ ヽ ヽ 巧 ヽ く く く と 占 宮 占 台 崎 由 ト 崎 ヽ い ゆ り い い 〕 ● 寺 日 ∞ 【 田 ６ 〕 Ｈ Ｎ 的 卜 寺 Ｈ 【 ヽ 【 「 Ｈ Ｈ 【 ド Ｈ Ｈ 代 ” Ｈ

W eek

FIG .3.Seasonal dcnsity ofD 閉夕切ル●夕れなわPθrr9わIII sugarcane fOr 1993,lowerRio Grande V alleyD T exas. M eans are fOr four flelds for each ofthe cultivars N C o 310 and C P 70-32 1. r'20 il. -E- Alamo 1.00 N C o 3 10 -^- Elsa 〕 Ｏ ０ ８０ Ｏ -o- ヨ \ LaFeria リ 出 -'- Ｏ ０ ６０ Mercedes ａ \

宮 電 ● 尊 宮 ０ ４０ \_-=--^

Ｌ 】 ｈ 」 」 ｈ 静 あ あ ヽ ぁ ロ Ｅ ロ 宮 【 Ｈ 【 【 的 的 的 Ａ ョ ョ ョ 事 ミ 車 ａ 才 阜 ａ 田 付 Ａ 電 ョ ョ ョ ヨ ヨ ヨ ヨ 専 ｇ く く 才 呂 日 Ｅ 耳 Ｈ 巧 巧 特 ヽ ヽ 特 恥 鳴 く く く 々 台 ぐ、 ヽ 宮 白 白 〕 ● 卜 や Ｈ ∞ 崎 Ｈ ∞ 寺 Ｈ ∞ 崎 的 Ｈ ∞ い 代 へ 〕 ” ● Ｈ Ｎ Ｎ 再 再 田 い Ｎ 日 Ｈ い 【 Ｎ 的 い Week t''o b. -tr- Elsa ‐ 1.00 -A- LaFeria C P 7 0 3 2 1 〕 ０ ０ ８ -o- ０ ０ Mercedes

ヨ

リ 出 -a- ● Runn ａ ０ ６０ 目

電 Ｏ Ｆ を ０ ４０

0.00 ０ ● Ｌ Ｌ Ｌ と 」 」 ｈ 」 ヽ あ ぁ ヽ あ 宮 宮 ロ 宮 【 「 一 【 的 ● Ｏ 電 車 電 田 亀 ａ ａ 中一 ” 車 毎 毎 電 ョ ョ ョ ョ 専 Ｈ ヨ 戸 ョ ヨ ヽ ヨ 』 Ｌ 専 専 く く く 専 噂 噂 日 日 車 ヽ 中 巧 ヽ ， ヽ 巧 を 占 岳 宮 ‘ 「 々 白 白 白 占 ト 「 寺 日 ∞ い ヽ ト 寺 Ｈ ∞ 寺 「 ∞ り 〕 ゆ 碕 ● 卜 ヽ い い い ６ 的 ● Ｎ 【 Ｈ ば Ｈ Ｈ い い Ｈ Ｈ Ｈ 田 ” Ｈ Week

FIG. 4. Seasonaldensity ofD raeculacephalaportola in sugarcanefor 1994,lower Rio Grande Valley,Texas. Meansare for four fields for eachof the cultivarsNCo 310 andcp 70-321. in 1993. For eachspecies, there were no differencesin Taylor a or Dbetween sugarcane cultivars (P > 0.15). Theseresults indicated that underTexas sugarcane growing conditions and for commonlyencountered densities, S. saccharivoralns amoreaggregated spatial pattern than the othertwo species,

TABLE 1. Taylor Coeffrcientsfor ThreeSpecies ofHomoptera in TexasSugarcane, 1993- 1994. Species Year a (+ SEM) , (+ SEM) Saccharo sydn e sac charivora 1993 4.r5 (1.07) 1.55(0.03)a 0.974 58 t994 3.16(1.12) 1.43(0.06)a 0.907 53 b P erkins iella saccharicida 1ee3 1.48(1.11) 1.13(0.04) O.sSl 34 1994 1.10(1.12) 1.02(0.05) 0.935 32 Draeculacepha la p ort ol a 1993 0.85(1.07) 0.95(0.03) 0.934 61 re94 0.87(l.u) o.es(0.06) 0.794 6l a, b, Slopesignificantly different from I .0,F-test, P < 0.01,0.001 , respectively. 600

500 り -!- SI.saccharivora やＯ Ｏ ４００ ヨ -L- リ P. saccharicida 虫 〇 】 ３００ -o- ● D. portola ０ 日 再 ２００ Ｚ

100

0 ０ ４ ５

Density per shoot

FIG. 5. Estimated number of shoots required to estimate Saccharosydne saccharivora, Perkinsiella saccharicida, and Draeculacephala portola densities in Texas sugarcane. Due to this aggregation,larger samplesizes would be neededto estimatemean number of S.saccharivorapershootthantheothertwoinsects.Forexample,thenumberofshootsrequired to obtain90olo confidence limits which are*20o/oof themean of 5.0S. saccharivora would be ca. lO0 shoots(Fig. 5). The resultspresented here provide the first within-field distribution sampting data foi these homopteran speciesin Texas sugarcaneand form a baseline of information for growers,consultants, and researchers.

ACKNOWLEDGMENT

An early version of the manuscript was reviewedby N. Epsky, USDA, ARS; J. Irvine' TexasA&M University; J. Sivinski, USDA, ARS; and L. T. Wilson, TexasA&M University. R. Saldafia,J. Huerta and A. Olivaresprovided technicalsupport.

LITERATURE CITED

Abbott,E. V., andJ. W. krgram. 1942, Transmissionof chloroticstreak of sugarcane by the leafhopperDra eculacephalaportola. Phytopathol.32: 99-100' Abbott,E. V., C. G. Hughes,and J. P. Martin. 1961. Chloroticstreak, p' 37L'387.In J.P. Martin,E. V. Abbott& C. G. Hughes[eds.] Sugar-CaneDiseases ofthe World. Elsevier, Amsterdam. Allsopp,P. G.,andR. M. Bull. 1990. Samplingdistributions and sequential samplingplans for Perkinsiella saccharicidalSrkaldy(Hemiptera:Delphacidae) and Tytthus spp.(Hemiptera: Miridae)on sugarcane.J. Econ.Entomol. 83:2284-2289. Chang,V.C.S., andA. K. Ota. 1978. Feedingactivities of Perkinsiellaleaftroppers and Fiji diseaseresistance ofsugarcane. J. Econ.Entomol. 7l:297-3Q0. Charpentier,L. J. 1970. The West Indian sugarcanefulgorid in Iouisiana SugarJ. 32: 58'60' Egan,B. T., C. C. Ryan,and R.I.B. Francki. 1989.Fiji disease,p. 263-287.In C. Ricaudet al. [eds.] Diseasesof sugarcane.Elsevier, Amsterdam, the Netherlands. Francki,R.I.B., andC. J. Grivell. 1972. Occurre,nceof similarparticles in Fiji diseasevirus- infectedsugax cane and insect vector cells. Virol. 48: 305-307. Hall, D. G. 1988.Insects and mites associated with sugarcanein Florida. FloridaEntomol. 71: 138-150. Hawkins,J. A., B. H. Wilson, C. L. Mondart,B. D' Nelson,R. A. Farlow,and P. E' Schilling. I 979. Leaftroppersand planthoppers in coastalBermudagrass: effect on yield andquality andconhol by harvestfrequency. J. Econ.Entomol. 7 2: l0l-104. Meagher,Jr., R. L., J. W. Smith,Jr., andK.J.R. Johnson. 1994. Insecticidalmanagement of Eoreunialoftini (Lqidoptera: Pyralidae)on Texassugarcane: a critical review. J. Econ. Entomol.87 t 1332-1344. Meagher,Jr., R. L., S. W. Wilson,R. S.Pfannenstiel, andR. G. Breene.1991. Documentation of two potentialinsect pests of southTexas sugarcane. Southwestern Entomol. 16: 365- 366. Meagher,Jr., R. L., S. W. Wilson, H. D. Blocker,R. V. Eckel,and R. S. Pfarurenstiel.1993. Homopteraassociated with sugarcanefields. Florida Entomol. 76: 508-514. Metcalfe,J. R. 1969. Studieson thebiology of the sugar-canepest Saccharosydne saccharivora (Westw.)(Hom., Delphacidae). Bull. Entomol.Res. 59: 393408. SAS Institute. 1996. SAS/STAT guidefor personalcomputers, version 6. 12 ed. SAS Institute, Cary,NC. Sosa, Jr., O. 1985. The sugarcanedelphacid, Perkinsiella saccharicida (Homoptera: Delphacidae),a sugarcanepest new to North America detectedin Florida. Florida Entomol.68: 357-360. Sosa,Jr., O., R. H. Cherry,and R. Nguyen. 1986. Seasonalabundance and temperature f Su arcm e d g elp h ttid ぃ Om O p tera : D cぃh ttid ac). E n 宙Юn . E n tom d 。15 : 1補 料 私 i べ 拶 楽 ど で 撤瑠漁淵

T a押 工n g h e 調 』 d Sttb 航 恥 O 鼠的 p ttd 航 o飛 . 靴 .紹 ユ繍 球 岳基 野 晋 軸 名 発 品 1品 iで ズ ' 棚 ヒ 1艦 皆盟 樹 糾 掛 肘 ⅦS6 19も 古鮮 品縦i淵総 諾 ば 、齢 観撤 撤 部 艦 WVilson,L .T . 1994 .E stim atin a g bundance,m p act,an d interactiOns am ong arthropodsin cotton ぺ 鞘 襴 工 鯛 て ジ 督 M 炊 fA u軌的 rrhyncha On ttg範 “ PЮα6ぬ血 chcn. 粒品 亀 .苫踏ご罷識1 著財罷戦

０ V O L .2 8 N O .1 S O U T H W E S 羽団組W E N T O M O L O C IS T ヽLAぷこ2003

RELEASE OF TNCHOGMMMA PRETIOSUMIN COTTON WITH A NOVELGROUND SPRAYER

Allen E. Knutson TexasA&M Researchand ExtensionCenter 17360Coit Road,Dallas, TX 75252

ABSTRACT

A mechanicalsystem for applying insect natural enemiesin a liquid suspensionand adhering the natural enemiesto foliage was evaluatedfor the releaseof Trichogramma pretiosutnRileyin cotton.Immersion ofhost eggs contairingT. pretioszre pupae in amixture of commercial carrier @ioCarrierrM)and water for up to four hours did not significantly reduceemergence ofadults from host eggs.However, following applicationofhost eggswith a novel sprayer,the BioSprayerfr, emergence of adtit T.prelioszzr was significantly reduced by 22-30o/o.Significantly more of the host eggswrth T. pretiosum pupae(7870) recovered fromthecotton canopy after applicationwith the BioSprayerrMwerepresent intheupperplant canopy and l4Yo and 8% were recoveredin the middle and lower portion ofthe canopy, respectively. The rate ofloss ofapplied host eggsfrom cotton leaveswas 13.6%per day and was attributedto weatheringof the adhesive,abrasion from leavesand insect predation.

INTRODUCTION

Trichogrammaspp. are massreared and releasedfor augmentativebiological control oflepidopterouspests ofcrops and forestsin many countries(Li-Ying 1994). Trichogramma arereleased manually in many of theseareas and the lack of an efficient mechanicalrelease methodis oneofseveralconstraints to the commercialdevelopmerfiof Trichogramma forpest control in the U.S. (King et al. 1985). Bouseand Morrison (1985)developed an aerial applicationsystem for T.pretiosum Riley which maintainedpharate adults inside parasitized Sitotrogacerealella (Olivier) eggsat 13oC until releasedfrom the airplane. Adult emergence occurredwithin four hoursofrelease. This methodreduced mortality ofpupaeresulting from prolongedexposure to high soil surfacetemperatures in cotton fields. However,the needfor aircraft to be equippedwith a refrigeration unit, the additional rearing costs to produce temperature-programmedeggs and a reductionin vigor in T. pretiosum following exposwe to cold havelimited commercialinterest in this method(Morrison et al. 1998,Stinner et al. 1974). rn Europe, aerial releaseof cardboardcapsules containing pupae of T. brassicae Bezdenkohasbeencommercially successful forcontrol ofOstrinianubilatis (Htbner) incorn, but this technologyis not currentlyavailable in the u.s. (Bigler 1994,Suh et al. 2000). Gardnerand Giles (l 997)discussed the benefits ofsprayer technology for releasingbiological control agentsand demonstratedthat submergenceof T.pretiosun pupaeinside E. hthniella (Zeller) eggsin water for up to threehours and passage throug[ a large-orificefan nozzledid not significantly reduceadult parasiteemergence. . The.USDA-Af;.Sdereloped a low pressure,liquid systemfor applying and adhering eggs of chrysoperla spp. to pecan hee foliage (Teddersand Blythe 1998)and this system hasbeen used to apply c. rufirabris eggstocotton (Knutson and redders 2002). This sprayer system(BioSprayerw) andadhesive carrier(BioCarrierru) w"t. "oi-"."iuttl'a"'"topeo -o marketedbv SmuckerManufacturing, tnc., Harrisb*g, bR 97446. Ai;;;;gh designedto apply chrysopideggs, the systemwas believedto have potentialf"; ttl'rg pupaeof kic,hogrammaspp. due to their smallsize and durability res,irrire n"- ii#irltection insioe thehost egg. The purpose ofthis study was to evaluatethe BiosprayerrMsystem for rereasingr pretiosum in cotton plants for_augmentative biological "otttroi or t"pioopi".*, pests. The specific objectives were to deterrnine the effeciof expos're t" ih" ;;;t solution and applicationthrough the spray systemon emergenceof Tiichogramro "autt, rro- pupaein hosteggs, the retentiontime ofparasitizedhosi eggson cottonleaves and the distribution of hosteggs in the cottoncanopy following applicatiJn.

METHODS AND MATERIALS

TheBiosprayerrM systemconsisted of a 37.9-riter tank pressurized to 0.7 poundsper squareinch by a l2-volt DC air compressorand a variablepressure control valve. The spray solution was metered through a plastic tube that terminatedinside the end of a two-inch flexible metalpipe. liT.J"r Air flow from a 44-ccvariable speed blower passed down the flexiblepipe and carriedthe spraysolution onto the cottonfoliage. The endof the flexible pipewas positioned l0-15 cm abovethe canopy of a'rowof cottonlAir speedfrom theblower was maintainedat 30-35mph which broke up the streamof solutionlnto largedrops and depositedthem in the canopy.The BioSprayeit"*us mountedbehind a tractor,powered by thetractor's l2-volt batterysystem and equipped with trro spraylines and air hosesallowing applicationof Trichogramma to two rows. Calibrationwns ,rccomplishedby a-djustingth! pump pressureand tractor speed. The spray solution consistedof a 1:8 ratio of a non-toxic solution marketedas BioCarrierru(SmuckerManufacturing, Hanisburg, OR) anddistilled water. TheBioCarrierru was designedto maintain the host eggsin solution andalso to adherethe pupaeto the foliage afterthe spray deposit dried. Thespray solution and host eggs containinjr'pretiosum pvpae were agitated with a large paddlecentered in the spraytank androtated-by an electric motor mountedin the top. The compressoralso bubbled air into the spraytank to maintaina uniform suspensionof host eggsin the liquid carriersolution. Eggsof Ephestialafiniella (zerer) parasitizedby 7. pretiosumwerepurchased from " "o.-o.iul insectary@eneficial Insectary, oak Run' cA). The number of parasitizedeggs in ten subsamplesof 0.003 g each was countedto determineapplication rates. Emergenceof T. pretiosumafter Immersionin spray solution. Host eggs, ca.250, conlainingT. pretiosum pupae, were placed in eachof40 one-dram(4.5 ml), screw-topgrass vials filled with the spraysolution (l:8 BioCarrierrMand distilledwater). Ten vials were selected atrandom afrer l,2,and 4 hrsand the solution was filtered through a Buchnerfunnel to recover pupae, the A subsampleof 25 pupaewas removed from thefilter paperwith a fing moistbrush and placed individually into cellsofa tissueculture plate and incubated at 7g.F and 80% R. H. After one week, host eggswere examinedto determineif an adult wasphad gmerged Each time period was replicatedten times with a replicateequal to a subsamileof 25 pupaeper vial. Host eggsnot immersedin the spraysolution r.*.d * a control. Emergenceafier Dischargefrom theBiosprayerru. To measurethe effect of passage throughthe Biosprayerru on adultemergence, two gallonsof spraysolution and ca 3g,000 r pretiosumptpaewere added to the spraytank of theBiosprayerrM. The largepaddles inside the tank provided continual agitation to keep the pupaein suspension. thi ,p.uyo *^

２ calibratedto deliver 160ml per minute at a pump pressureof 0.7 psi anda blower speedof 3l mph. The spray stream from the BioSprayerru was collected into a 500 ml-graduated cylinder. T. pretiosum pupaewere rernovedfrom the spray solution by filtering it thiough a Buchner fururel. Pupaewere removedfrom the filter paperwith a fine brush and placed in individual cells ofa tissueculture plate. Fotr subsamplesof40-50 host eggswere collected andheld for adult emergenceas described above. Host eggsnot exposedto the spraysolution servedas the control. The.experimentwas repeatedon threedates. Distribution of Host Eggsin the CottonCanopy. Field experimentswere conductedto determinethe distribution of host eggs in the cotton canopy when applied with the BioSprayerrM.White, paperindex cards(7.8 X 12.7cm) wereplaced at 25, 50 and 75 cm abovethe soil in the cotton canopy. The averageplant heightwas 80 cmoso the cardat 75 cm was placedon a fully expandedleaf in the plant terminal. Host eggsapplied with the BioSprayerrMadhered to the cards and were readily visible for counting. Cards were positionedflat againstthe upper leaf surfaceand stapledto the leaf at four corners. A single cardwas placed at thethree heights on eachoffive plantsin thecenter ofeach often 90 foot rows of cottonrivithin a O.5-acreplot. The plot wasplanted to 'Delta Pine50'cotton on 40- inch row spacingsat the TexasA&M University Researchand Extension Center, Dallas, TX. Pupaewere applied at a high rate, 1,6 million per acre,to increasethe recoveryrate of eggs on the cards. The BioSprayerrMwas equippedwith two spraytubee, each centered above a single row. Each spraytube was I 0- I 5 cm abovethe centerof the cotton row and directed backata45"angletothetopofthecanopy. Tractorspeedwas3.4mph,airspeedexitingthe spraytube was 35 mph,and the spray output was 80 ml per 100feet of row, or 2.4gallons per acre. Pupaewere appliedto the cottoncanopy on I I August 1995. Air temperaturewas 95 "F andwind speedwas 5 mph. Fifteen minutesafter application,all ofthe cardswere recovered and the numberofhost eggsper card was recorded. Retention of Trichogramma Pupae on Cotton Leaves. Host eggs containing Trichogrammapupae were applied with the BioSprayerruon 9 August 1995 as described 'Delta aboveto a O.S-acreblock of Pine 50' cotton plantedat the TexasA&M Researchand Extension Center at Dallas, TX. Immediately after application, the terminal foliage was searchedfor host eggs. Oncean eggwas found, a small arrow was drawnon the leafpointing to the egg and a red ribbon wastied to the leafpetiole andnumbered to assistin relocatingthe egg. Each egg was examineddaily for six days. The experimentwas replicatedtwice with eachreplicate a single row eachwith 33 markedpupae. In the secondexperiment, host eggs were applied through the BioSprayerrM as described aboveand were collected in a petri plate. Eggswith a Trichogrammapvpae were removed from the spray solution with a fine brush and placedindividually on the upper surfaceof a terminal cotton leaf. Spray solution on the egg adheredthe egg to the cotton leaf. A circle wasdrawn around each egg with a markingpen to identifuthe position of eachegg on theleaf. Thepetiole of the leafwas circled with a 4-cmband of TanglefootrM(Tanglefoot Company, GrandRapids MI) to preventaccess to the eggsby fire antsand otherpredators. Eggswere examinedl, 3, 4 and5 daysafterplacement on the leaf. The experimentwas replicated five timeswith eachreplicate a singleleaf with 25 pupae.Tests were conducted in mid-August when maximum temperatureswere 95-l00oF. No rainfall or dew occurredduring the experiments. Dataonpercentageemergenceof adullTrichogrammafromhosteggsweretransformed by arcsinbefore analysisofvariance and unhansformedvalues are reported. Differencesin "/' mean values were comparedusing Student's test for paired treatmentcomparisons and Fisher'sLeast Significant Difference (LSD) at a = 0.05. RESULTSAND DISCUSSION

Emergenceof T.pretiosum after Immersionin SpraySolution. The meanand standard error for emergenceof adurl T.pretiosum from pupaenot immersedin the Biocarrieril was 973%+0'9' Thesevalues- forpercentageemergence forpupae immersed in theBioCarrierru for 1,2 and4 hourswere92.96/o+2J,b0.9%iz.A g2.4%+ _o 1.7,respectively, and were not sigaificantlydifferent from emergencein the control(F= r.7r, d. fl : 3,36,p = o.rg). Resultssuggestthat immersionofhostiggs c ontaiungT.pretiosrr. p,rpu"irrtrr" BiocarrierrM for up to four hoursdid not significantly-reduce percentageemergence of adults. Morrison et al. (1998) concludedthat the emergenceof T. pretiosun rouowing immersionin the BiocarrierrMfor three hourswas similar to that foilowing immersionln water. However, emergencewas reduced by 6-l l% following immersionfoi 1-3hours, respectively, and was sigrrificantly different from that in the non-immersedcontrol. Emergenceafter Discharge from the BiosprayefM. Mean percentageemergence of adults from host eggspassing through the BioSprayerruwas sigrrificanttf t"s, trran ao..lt emergencefrom the control group orl eachof the threesample datJs flable i). Theseresults indicatethat some T.pretiosum_pupae sufferedinjury dwing agitationin the spraytank and passagethrough theBiosprayerrM, resulting in a decreaseof21-io% in subsequentemergence of-adults from host eggs. Modifications to the tubing alignmentand tubing attachmentto solenoids andcouplers have since been made to reduceconstrictions to the flo:wand potential injury to the pupae @. Morrison, personalcommunication). Thesestudies will needto be repeatedon the modified BioSprayerruto determinethe need,if any, for increasesin release rates to compensate for reduced adult emergenceresulting from passagethrough the BioSprayerrM.

TABLE I . PercentageEmergence (Mean+ S.E.) ofAdultT. pretiosumFollowingApplication of ParasitizedHost EggsWith the BioSoraverru. C ottrolと ｄ ｉ BioSprayerre P ‐v alu e

20 Ju ly 81±3b 59±5 a ６ 14 .9 0.008

4 A u gu st 88 ±l b 6 6 圭2 a ２ 177 ,4 0.006

8 A u u st ± ６ g 86 2b 56±7a 4 9 .3 0,004 fo ll ow ed sameletter a ro w are n o t s different; t- test, a =

Distribution of Host Eggs in the cotton Canopy. A total of 339 parasitizedhost eggs were recoveredfrom the cardspositioned in the cotton canopy(Table 2j. Significantly moie of the recoveredeggs, 78Yo, were found in the top canopythan at the two lJwer leveis (F = II'6'd. f.=2, 147,P=0.00r).Also,themeannumberof eggspercardwassigrificantly greater in thetop canopythan in the middleor bottomcanopy (F = 1L63, d,.f. = 2, 147,p : 0'001) and the number of cardswith one or more eggswas significantly greaterin the top canopythan in thebottom canopy (F : 6.14,d. f. : 2,27, p :0.006). Theieresults suggest the majority (78%) of the host eggsrecovered from the canopywere placed in the terminalor upper25 cm of thecotton plant. Variablessuch as height of thecotton plant, leaf surface area andblower speed could potentially influence the relative distribution ofpupae in thecanopy but werenot evaluatedin thesetrails. RetentionofApplied Pupaeon CottonLeaves. Retention ofhost eggson cottonfoliage over time when appliedwith the BioSprayerrMwas describedby the relationshipr= gg.67_

４ TABLE 2. Distribution ofParasitizedHost Eggsin the CottonCanopy Following Application with the BioSprayerru. Canopy Total eggsrecovered Mean number Mean numberof cards zone (percentage) eggspercard" wittl >legd

T op 264 (78) a 5.3 a 2.8a

M id d ic 48 (14) b O.9b l.8 ab

B ottom 27 (8) b O.5b l.lb uMeans followed by sameletter within a columnare not significantlydifferent. LSD, c : 0.05

13.64X where Iis the percentageof.pupae present andXis the numberof days after application(Fig.l).Thisrelationshippredictedalossrateof13.6%ofthehosteggsperday.

２０

や ００ Ｅ Ｏ A の ０ ―― 」 ００ 住 Ⅲ ■圭_ 占 ０ 『 ａ 側 コ 住 一Ｅ ４０ Ｏ 翌 ｏ Y = 98.4- 3.46X 住 郷 f = O.29 ０

００

９０ ど ０ ８０ り Y = 38 .67 ‐13 .6 X ｐ ７０ヽ 住” ｏ ６０ ｃ ａ ５０ ョ 住 ４０ 一ｃ ３０ Ｑ せ ２０ ０ 住 １０

０

3 4

D ays A fte r 押, p licatiO n

FIG. 1. Relationshipbetween retention ofparasitized host eggs on cottonleaves and days after applicationwith the BioSprayerruand with preditors excluded(A) or predatorswith access to host eggson leaves@). when predatorswere excludedfrom accessto the eggs,this relationshipwas r: 9g.4 - 3.46 xand the rate of loss of eggsfrom the leaf was onry3.5%per day Gig.r;. Theseresults suggest that removal of host eggs by predators was an important source of pupal disappearance. The red imported fire ant, solenopsisinvicta Brren preys on lepidopteran eggsin the cotton canopyand was commonin the stuOynetcl. tn earlieitrials in this field, fire ants were observed to remove up to 50% of the E kuehniella eggs parasitizedby Trichogramma within 24 hourswhen eggswere glued to cardsand placedli cottonterminals. In addition to predation, variable wind effects between experiments could also have contributedto the differencesin rateof lossofhost eees, No rainfall or dew occurredduring thesestudies, both ofwhich would be expectedto increaseloss ofpupae asthe BioCarrierru is watersoluble. Leavesmoving acrossthe surface of otherleaves during windy periods may havedislodged some of the o*";. Removalof pupae by fire ants and other predators may also have been importani. These results demonstratethat applicationofpupae shouldbi timedsuch thatthe T)ichogramma a&iltswiLl emergewithin -2 I daysfollowing release. Also, release rates can be increasedto compensate for the expectedloss ifdetachment and predationrates are known. These evaluations. indicate the BioSprayerru system can eflicienfly apply Trichogramma pupae inside host eggs to the cottin canopy. while there is somepupal mortality due to exposureto the Biocarrierru solutionandpassage through the spraysysrem, the application rate ofhost eggscan be increasedto compensatefor thii lossand therefore achievethe desiredrelease rate of rrichogramna adults. The BioSprayerrMdeposits a majorityofthe.host eggsin the uppercanopy where Heliothine eggs,a 'Efhcacypoiential target pest for augmentation of rrichogramma. are typically depositedl of released Trichogramma would presumablybe increasedby placingpupae close to thetarget pest; thus reducing time fcr the parasiteto searchfor the host. Atso, aitrering the pupaeio the foliage avoids exposure of pupae to fatally high temperatureson the soil surface. Releaseof Trichogramma shouldbe timed to ensureadult emergenceoccurs within one to two daysof releaseto minimizeloss of pupaedue to predationand detachment from the foliage.

ACKNOWLEDGMENT

t This work supportedby a grant from the CooperativeState Research, Education and Extensionservice, project 93-EpMo-l-0400. Appreciationis expressedto charlesSuh and Louis Teddersfor reviewing a draft of this manuscript.

LITERATURE CITED

Bouse,L. F., andR. K, Monison. 1985. Transport,storage and release of Trichogranma pretiosum.Southwest. Entomol. Suppl. 8:36-48. GardnerJ. and K. Giles. 1997. Mechanicaldistribution of chrysoperlarufilabris and Trichogrammapretiosumi survival anduniformity of disclrargeafter spraydispersal in an aqueoussuspension. Biol. Control.8:138-142. lQng'b'.G.,K.R.HopperandJ.E.powell.19g5. Analysis-ofsystemsforbiologicalconfrol of croparthropod pestsin theU.S. bjr augmentati-onof iireaator. *Jprrllt"s, p. 210_ 227InM-HoyandD.Herzog [eds.].BiologicalcontrolinAgriculturallpMSystems. AcademicPress. Knutson,A. E. and L. Tedders. 2002. Augmentationof greenlacewing, Chrysoperla rufilabris, in cotton. Southwest.Entomol. 27 :231_239.

６ 町 鞘 縄 辮 岳慾 欝 熙 盟 糾 温 謎艦

鋭mer 舶予酔捕撤 崩払換艦 古!稚e総岳辞盤鮒艦 笛盤 560. Stt C α 品 と込獄 キ! え鮒登 鋪 鑑 開 iip diil骨告♀;【格螢告筈盤 縄 、 Tedd 駐Benefldd hstt e s 鴫,粘,柑品最盤鰐 gg pり噌d針たo Patent m . V O L .2 8 N O .1 S O U T H W E ST E R N E N T O M O L O G IS T M A R .2003

ARTIFICIAL MATURATION OF FEMALE ALATES FORTHE PURPOSEOF THE PRODUCTIONOF ONLYMALE SOLENOPS$INVICTA (HYMENOPTERA: FORMICIDAE)

BashirMirr, S. BradleighVinson2 and Jorge A. Piedrahital ABSTRACT Hereinwe describea methodof artificial matwationof l-3 - day old, post-eclosion female imported fire ants,Solenopsis invicta Bwen, tluough the use of virgin queensatellite colciniesmaintained by a parentcolony that producesonly haploideggs and malebrood on a sustainedbasis. Productionof largenumbers of sucheggs with a predictableoutcome will provideconsistent material that can be usedto focuson malebiology and allow for various geneticmanipulations that assistthe productionoftransgenic fire ants.

INTRODUCTION

Ant colonies primarily consist of females that include queen(s),workers that are usually sterile, and female alatesthat are reproductivelyinactive; malesare ganerallypresent for a limited periodwithin a colony (Wilson 1975). The occurrenceof malesis regulatedin part by the queen and in part by the workers resulting in queen-workerconflict over sex allocation(Trivers and Hare 1976)and is a subjectofgreat interest(Bourk andFranks 1995). Brian (1969) showedthat workersof Myrmica rubraL. laid haploid eggs,but theseeggs often serveas food for developingqueen-laid eggs, although he was able to rear malii. Passeraet al. (1988) working with the Argentine arfi, Linepithemahunile formally Iridomyrmex humilis (Mayr), reported that queenslaid male eggs throughout the year, but most failed to reachthe pupal stage.This failure was influencedby the ratio of workersto larvae.If therewas plenty of proteinavailable and a high worker to larva ratio, moremales reachedthe adultstage (Passera et al. 1988). Importedfire ant, SolenopsisinvictaBtxen (Hymenoptera:Formicidae) workers are sterileand reproductionis by the queen. This processbegins following a matingflight that result in insemination,dealation, and her transformationinto a queenthat produceseggs for the remainderofher life. The eggsshe produces have different fates: haploid, non-fertilized eggsdevelop into males; diploid, fertilized eggsdevelop either into fertiie femalesor sterile workers;and trophic (non-embryonated)eggs are usedas food. A colony is alwaysin a dynamicstate consisting of different classesof progenysuch as workers,males, and alate femalesin variouspioportions at differenttimes dependingon a numberof environmental conditions(Vargo and Fletcher1986, 1987). However,a majority (about97o/o) of queens, eggsdevelop into sterileworkers. Lessthan 3-4% developinto fertile malesand females. However,in polygynefire ant colonies,some queens produce diploid maleswith various

Departmentof Veterinary Anatomy and Public Health.Texas A&M University,College Station, TX. 77843, USA. 2 Departmentof Entomology,Texas A&M university,college Station,TX. 77g43,usA.

９ degreesof sterility (Hung and Vinson 1976,Ross and Fletcher19g6). As a result,males are not consistentlyproduced and not all malesproduced are fertile. Geneticmanipulation of the fire ant will require consistentproduction of a large numberof--reproductively competent adults that can be manipulatedgenetically and still generateoffspring. The preferredapproach is to geneticallymanipulateihe egg stageofthe reproductives'Unfortunately, the socialnature ofthe ant colonypresents obstacles ior using fot. genetic 9p^g^t manipulation.These obstaclesinclude an inability to non-destructivel! diffe.rentiate betweenhaploid, diploid, or trophiceggs immediately after oviposition-female,and the inability to predictwhether a fertilizeddiploid eggwill developinio a fertile a sterile worker,or a sterilemale. Moreover,manipulation of eggsby biolisticsor microinjectioncan result in significant mortality. Thus, for efficient production of transgenicS. iniicta, large numbers.ofl-2-h old eggsthat can be manipulatedto developinto ieproductiveforms of propagationof a transgeneare essential.Also it will be importantthat all eggsconsistently have one fate. Becausethe fate of fire ant diploid .ggt *i influenced by lxternal factors (Robeauand Vinson 1976),or geneticfactors in the caseofsterile males(iloss and Fletcher 1985),diploid eggsdo not appearto be a good option for manipulation.i-Iaploid eggs may offer a solution. In a normal fire ant colony, the queeninfluences the overall behaviorof the colony with a batteryof pheromones(Vargo 1999,obin and vander Meer l9g9). some of these pheromonesprevent dealationand the reproductivedevelopment of virgin female alates (Fletcherand Blum 1981,Sorensen et al. 1985,virgo and Forter 1993,iargo and Laurel 1994).Both mechanicallydealated virgin femalealates and virgin female alatesremoved from the inhibitoryinfluence of a queenpheromone will initiateovary development (Glancey et al. 1981, Fletcheret al. 1983). But, virgin female alatesdealate more readily in the presenceof workers(Fletcher et al. 1983)and would presumablyinitiate ovary development and lay haploid eggs sooner.However, these studieshave been limited to addrissine questionsconcerning the attractionof workers, dealation,wing muscle histolysis,ovar! !e_v.elogm9nt,and egg laying by virgin female alates.Stringerl al. (1976) reportedthat individual dealatesfrom multiple queencolonies were incapableof founding coloniesand died within 12 wk when separated;however, these authorsfound that dealatesurvival was enhancedwith the addition of workers. Further, 98%oof over-wintered and spring-reared virgin queens that were placed separatelyin laboratory rearing conditions initiated ovipositionand colony-foundingbehavior, but only 73%odealated. Their eggswere larger than normal and less than l%o were viable (Fletcherand Blum l9g3). F-i;ld rnonogfrr" coloniesof S' invicta orphanedby removal of a functional queendevelop ieplacement q.iens from dealates.alreadYpresent in the colony. A proportionbfthes" replacementqueens were not inseminatedand producedonly male brood (Tschinkeland Howard l97g). Although male pupae were observed,no male production was reported.None of these studGs demonstratethe production of adult males, and no studies have maintained queenswith a goal of producingnumerous males or only producingmales. Further,such colonies were not sustainedbecause workers were not replaced. There have beena number of studiesusing virgin female alatesto determinewhether lactors'such asjuvenile hormone (JH) or carbondioxide, initiate dealation, subsequent ovary development,and egg production(Backer 1979; Fletcher and Blum lggl; l9g3; Fletcheret al. 1983;Keamey et al.1977;Jones et al. 1978).But again,the goal ofthese studieswas not maleproduction. Since virgin femalesproduce only haploideggi, we examinedthe potential to use such fernalesto produceadult males.We describea methodof artificial maturationof l-3-day old post-eclosionvirgin femalealates without matingand establishmentof satellite coloniesthat produceonly haploideggs, and, hence, only fertile maleson a sustainedbasis. The eggs from such satellitecolonies can be pooled together,and a foreign genecan be introducedby variousmethods to producetransgenic males that may be usedio propagatethe

０ foreign geneby natural mating. Altemately, the production of males will also allow for the planningof studiesthat canfocus on malebiology.

MATERIALS AND METHODS

Field colonieswere excavatedwith shovels,hansferred to bucketspreviously coated insidewith Talc (Sigma,St. Louis, MO) to retainthe ants,transported to the laboratoryand allowed2 daysto reestablisha tunnelsystem by the workers.Colonies were separatedfrom soil by floatingout the antsand brood (Jouvenaz et al. 1977);ants and brood were transferred with a spaghettistrainer to a plastic shoebox(27x40x9cm) coated with ADI Fluon (lCI Fluropolymers,Bayonne, NJ) alongthe top edgesto preventant escape.The box contained two Castone-floor(Densply, York, PA) petri dishes(l50xl5mm) that servedas artificial nests. Eachnest had a l5mm hole throughthe lid to allow queensand workersaccess. One filter paper(150mm diameter) was placedon the top of eachnest chamberto reducelight disturbanceto ants. Theselaboratory colonies (refened to a parent colonies)were housedin an environmentalroom at 28*1.2"C, 44-66% RH, and a photoperiodof 14:10 (L:D). Colonieswere maintainedon a daily diet of water,25Yo honey water, yellow mealworms (Tenebriomolitor L.) andartificial diet (Kuriachanand Vinson 2000) as needed. Maturationof virgin femalealates by variouschemical and physicaltreatments was evaluatedby a seriesof testsin which ten small nestswith 30 workerswere set up. These nestsconsisted of a Castone-floorpetri dish (95xl5mm) with a 15mmhole throughthe lid that servedas an artificial nest.These were placed in a shoeboxand maintainedas described above.In eachcase, a treatmentfemale alate was placed in eachnest and monitored daily for lossofwings, occurrenceofegg-laying, and subsequent embryonic development. There were four female alate treatments.A juvenile hormone treatment evaluated the effect on egg productionofthe JH analog(methoprene) in acetone(l00ng/ul) appliedto l-3-day old, post- eclosionvirgin femalealates removed from a polygynecolony. Five ul was appliedto the bottom of a 3ml plastic vial to which the female alate was introducedone hour later. After three days, the JH exposedfemale was placed in one of the small 30 worker nests. In a secondtreatment, one to threeday-old, post-eclosion virgin femalealates were placed into a chamberand exposedto 100%carbon dioxide (COz) for l0 minutes.The chamberconsisted of a 60mmplastic petri dish connectedto a carbonCOz via plastictubing. After recovering from COz anesthesia,the COz exposedfemales were transfenedindividually to 3ml plastic vials. After threedays, these treated females were placed individually into one of the small 30 worker nests.A third test consistedof l-3-day old, post-eclosionvirgin female alates placedin a 60mm petri dish andtransfened to a refrigeratorat 4oC.After 48h, the petri dish wasremoved from the refrigerator.Females were transferred individually to 3ml plasticvials and held at room temperature.One day later, treatedfemales were placedindividually into one ofthe small 30 worker nests.A control consistedofa 1-3-dayold, post-eclosionvirgin femalealate removed from a polygynecolony, held oneday, and thenplaced into oneofthe small30 workernests. One additionaltest consistedof placing virgin female alatesinto large queenless satellitecolonies to increaseaccess to resourcesthrough workers from a parent satellite colony.These colonies were initially setup with two virgin femalealates (ages I -3-daypost- eclosion)removed from a largecolony and placed in separatenests (l50x15mm as described above)with about 1000 workers from their parent colony. These satellitecolonies were maintainedfor 15 days and then the dominantfemale, the female producingthe greatest numberof eggsand attractingthe greatestnumber of workers(Chen and Vinson 2000),was determinedand retained.The other female was removed.Because these satellite colonies were unableto replenishlost workers,workers were addedfrom the parentcolony on a weeklybasis to maintainapproximately 1,000 workers per queen.These parent and satellite

２ colonieswere housed in an incubatorat27oc,50% RH, anda photoperiodof 14:10(L:D). colonieswere maintained on a daily diet of water,25%ohoney water, yellow mealworms(z molitor) andfrozen crickets. Four satellitecolonies were set up from both monogyne(single queen)and polygyne (multiple queen) colonies (8 total) andwere maintained over 6 months.

RESULTSAND DISCUSSION

All treatmentswere successfulin inducingwing loss and egg laying, but few eggs hatchedand none reached the pupal stage.Methoprene accelerated wing lossand egglaying comparedto other treatments,but there was no further developmentof theseeggs. Mosi likely theseeggs were trophic (Glancey et al. 1973,voss 1981, Vossand Blum I 9g7). Both the conhol groupand the satellitecolonies were slower to dealate(Table l) andbegan to lay eggsat aboutthe sametime. Althoughsome of the eggsin the controlgroup hatched, none of theselarvae reached second instar and the queen'soviposition rate declined.

TABLE 1: Effect of various Treatmentson the Maturationof virgin FemaleFire Ant Alates.

D av s after tream e耐

６０ Treatments 5 10 15

・ Ａ＋Ｂ Ｂ Ｂ Ｂ Control 10 Ｂ Ｂ Ｂ Ｂ Ｂ Methoprene A 10 Ａ Ｂ Ｂ Ｂ Ｂ Carbon dioxide 10 Ａ Ｂ Ｂ Ｂ Ｂ 4"C exposure 10 ・ Ａ＋Ｂ Ｃ Ｄ Satellite colonies E ttF 8 "Letters in the columns (A=Lost wings; B=Egg-laying;c:Egg devetopmenqo=tarvae; E:Pupae; F:Adult.) refer to the first appearanceof the characteristicfound in a majority of the nests.

In contrast,queens in satellitecolonies had larvae by 30 days. Further,with one exception,larvae were continuallyfound and,by 60 days,pupae and someadult maleswere found (Table1). We foundthat maleproduction increased over time, increasingthree to four timesover the third monthcompared to the first two months(Table 2). Colony#21g5(b) was removedfrom the studyafter 60 daysas this femaledealated but neverlaid eggsfor reasons that arenot clear.By the fifth and sixth month(data not shown),production in the remaining nestsincreased by another1.5X, but then leveledoff(data not shown)to an approximate averageof 100 to 200 males/month.There were no differencesin male productionusing virgin femalealates from eitherpolygyne or monogynecolonies. At the end of six months, we werestill successfulin addingworkers from the parentcolony to the satellitecolony that continued to produce males for several additional months at which time thev were terminated.

２２ T A B L E 2 1 M alc P rodu ction by S atcll■e F ire A nt C olon ics.

P arent C olony S atelllte No. malesproduced co lonv tvDe C0 10nv O-60 davs 62 ‐120 d avs 12 1-180 a d avs T otal

３０ #2 504 P ol ne a 120 170 320 ygy ３７ b 129 205 37 1 ２８ 浮2400 P ol nc a 151 2 12 39 1 ygy ３９ b 113 207 359 ４‐ 祥2 509 M ono ne a 105 16 1 307 gy ３２ b 137 2 10 379 ４５ #2 185 M onogyne a 114 131 290

M enn + ミD a 36 ±6 124 ±15 185±3 1 34 5±39 " Statisticalanalysis by ANOVA showsa significantincrease in male productionover time (P<0.001)

Removalof virgin femalealates from coloniesand stimulatingdealation, subsequent ovarial maturation,and ovipositiondid not produceviable eggs.Such queensmay not be physiologicallyprepared to becomereproductively competent. Virgin femalealates, under the influenceof a residentqueen and environmentalfactors, become sexually mature and wait for an environmentaltrigger that will initiate a mating flight. Environmentaltriggers (Monil 1974,McCluskey and McCluskey1984, Hooper and Rust 1998)initiate changesin the behavior of the membersof the colony, more specifically in the female reproductives (Obin andVander Meer 1994,Alonso andVander Meer 1997).Vinson et al. (1980)reported that prior to the mating flight, lipids in the digestivesystem of virgin female alatesincreased, suggestingthat females were stimulatedto maximize their nutritional reservesjust prior to their mating flight. As a result, we speculatethat virgin female alatesremoved from a colony prior to exposureto environmentalconditions that trigger a mating flight do not have the nutritional resourcesneeded for production of viable eggs. As a result, viable eggs were never produced by any of the treatments applied to the virgin female alates or control. Although removal of the virgin female alates from the queen's dealation and reproductive inhibitory pheromones(Obin et al. 1988; Vargo 1999) is an important step in initiating reproduction,another factor is providingnutrition. As shownby Sorensonet al. (1981,1983), laryae are important to the nutritional needs of the colony through the digestion and processingofproteins. Thesenutritional resources are transferred through the workersto the queen.Virgin femalealates in controlnests, once separated from the parentcolony, not only lackedthe resources,but alsoprobably received little additionalnutritional support from the limited number of workers (iust 30) that were added.This problem was overcomeby providingvirgin femalealates with approximately1,000 workers that collectivelyprovided the neededresources, some of which may be importantto the early developmentof larvae. Furthermale productionincreased with time in thesesatellite virgin queencolonies. This may be due to additionalnutritional resources provided by the presenceof the male brood itself andby the weekly additionof workersfrom the parentcolony. It is not clearwhy one satellitecolony failed to produceeggs. The satellitemethod of male productionwas not only successfirl,but may provide somebenefits. One advantageincludes agcess to a large numberof virgin female alatesover an extendedperiod of time that can be used to form a number of male producing satellite colonies supportedby one parent colony. Another benefit is the opportunity to maintain the male-only production over a long period of time through periodic worker addition. This ntStttOn伽h art tmug品業強鞘淵押患悠慧魚 堪館群総i d p ,

L IT E R A T LIR E C IT E D 畑伽的 ° 品挽 祐 撤 軽F 糖 鞠 灘 鮮品 盟ぽ i開 撤 鮮

Fletche n on de絲 鴨 辮 鑑 ,名規牌 鞘 群描 電艦 F珊 軸抗e 価°rS i tti尾ポt&とP縄 !縄批s盟撤;協呈糖 nnw耐唾deal愉町

鴨 鱗 麒 ギ 騨 告愚 茂齢 艦

２４ Mccluskey,E. S., andD. K. Mccluskey. 1984.Hour of matingflight in threespecies of ants (Hymenoptera).Pan-Pacifi c Entomologist.60: I 5I - I 54. Morril, W. L. 1974.Production and flight of alatered importedfire ants.Environ. Entomol. 3:265-271. obin' M' s., B. M. Glancey,w. A. Banks,and R. K. vander Meer. 1988.eueen pheromone production and its physiological correlatesin fire ant queens (Hymenoptera: Formicidae)treated with fenoxycarb.Ann. Entomol. Soc. Amer. gl:g0g-g15. Obin, M' S., and R. K. VanderMeer. 1989.Nest mate recognition in fire ants.Ethol. 80:255- 264. Obin, M. S., and R' K. VanderMeer. 1994.Alate semiochemicalsrelease worker behavior duringfire antnuptial flights. J. Entomol.Sci. 29:143-151. Passera,L., L. Keller, 41fl J. $rzz61i. 1988.Control brood male production in the argentine antIridomyrmex humilis (Mayr). InsectesSociaux. 35:19-33. Robeau,M. R., and S. B. vinson. 1976. Effects of juvenile hormoneanalogues on cast differentiation in the imported fire ant, Solenopsisinvicta. J. Georgia Entomol. Soc. 3:198-203. Ross,K. G., and D. J. c. Fletcher.1985. Genetic origin of the male diploidy in the fire ant, Solenopsisinvicta (Hymenoptera:Formicidae) and its evolutionary significance. Evolution.39:888-903. Ross, K. J., and D. J. c. Fletcher. 1986. Diploid male production-asignificant colony mortalityfactor in the fire anlSolenopsis iniicta (Hymenoptera:Formicidae). Behav. Ecol.Sociobiol. 19 :283 -29 l. Sorensen,A. A., J. T. Mirenda,and S. B. vinson. 1981.Food exchange and distributionby three functional worker groups of the imported fire ant, Solmopsis invicta Bvren. InsectesSociaux. 28:383-394. Sorensen,A. A., R. S. Kamas,and S. B. vinson. 1983.The influenceof oral secretionsfrom larvae on levels of proteinasesin colony members of Solenopsisinvicta Bwen (Hymenoptera:Formicidae). J. Insectphysiol. 29 163-1 69. Sorensen,A. A., D. J. c. Fletcher,and s. B. vinson. 1985.Distribution of inhibitoryqueen pheromoneamong virgin queensofan ant,solenopsis invicta. psyche. 92:57-69. Stringer,c. E. Jr., w. A. Banks,B. M. Glancey,and c. S. Lofgren. t976. ted importedfire ants: capability of queensfrom establishedcolonies and newly mated queensto establishcolonies in thelaboratory.Ann. Entomol. soc. Amer. 69:i004-1006. Trivers, R. L., and H. Hare. 1976.Haplodiploidy and the evolutionof the social insects. Sciencel9l:249-263. Tschinkel,w. R., and D. F. Howard. 1978.eueen replacementin orphanedcolonies of the fire ant,Solenopsis invicta. Behav. Ecol. Sociobiol.3:297-310. vargo, E. L. 1999,Reproductive development and ontogenyof queenpheromone production in the fire antSolenopsis invicta. physiol. Entomol. Zq:li\-216. vargo, E. L., and D. J. c. Fletcher.1986. Evidence of pheromonalqueen control over the production of male and female sexualsin the firi ant, solenipsis irwicta. J. comp. Physiol.159:741-749. vargo, E. L', and D. J. c. Fletcher.1982. Effect of queennumber on the productionof sexualsin naturalpopulations of the fire ant, Solenopsisinvicta. physiol. Entomol. l2:109-l16. vargo, E. L., and s. D. Porter. 1993.Reproduction by virgin queenfire ants in queenless colonies; comparative study of three taxa (soleiopsis richteri, tryurio s invicta/richteri,S. geminate)(Hymenoptera: Formicidae). Insectes Sociaux 40:2g3- 293. Vargo, E. L., and M. Laurel. 1994. studies on the mode of action of a queenprimer pheromoneofthe fire ant,sorenopsisinvicta.J. Insect.physiol. 40:601-610. Vinson,S. B., Phillips, S. A., and williams, H. J. 1990.The functionof the post-pharyngeal glands of the red imported fire ant, solenopsisinvicta Buren J. Inseit physiol. 26:645-650. Voss,S. H. 1981.Trophic egg production in virgin fire ant queeos.J. GeorgiaEntomol. Soc. 16:437-440. Voss,S., andM. Blum. 1987.Trophic and embryonated egg production in foundingcolonies of the fire antSolenopsis invicta (Hymenoptera: Formicidae). Sociobiol. 13:271-27g. wilson, E. o. 1975. Sociobiology:the new synthesis.Belkna press,Harvard University, Cambridge,MA. 697p.

６ V O L .2 8 N O .1 S O U T H W E S T E R N E N T O M O L O G IS T M A R .200 3

RELATEDNESSAMONG CO.EXISTING QUEENSWITHIN POLYGYNE COLONIES OF A TEXAS POPULATION OF THE FIRE ANT. SOLENOPS$ INI/ICTA

Y. P. Chen,L. Y. Lu2,L. C. Sko#, andS, B. Vinson3

lnsectBiocontrol Laboratory, PSVARSTSDA, Beltsville, MD 20705

ABSTRACT

Five microsatellite markers were isolated, characteized, and used as genetic markers to estimate the genetic relatednessamong fire wrt, Solenopsisinvicta Buren, queenswithin polygynous coloniesand amongqueens within aggregatedgroups in colonies maintained for a short time in the laboratory. Estimates of genetic relatednessusing microsatellitemarkers showed that co-existingqueens in the samecolonies were not closely related, which is in agreementwith previous findings using mitochondrial and nuclear markers. Queensalso randomly aggregatedinto groups, and queen aggregationwas not based on their relatedness. The F,1-valuesuggested that polygyrous colonies from the BrazosValley, Texas, can be consideredas a single population and demonstatedthat gene flow occurs frequently in this are4 probably by dispersion caused by mating flights, transportationand wind. The near zero Fis value indicatedthat polygynous queensin this areawere randomly mating.

INTRODUCTION

Polygyny is characterizedby coexistenceofmore than one inseminated,egg-laying fernale in a colony. Polygyny within a colony can be primary or secondary. primary polygyny ariseswhen severalqueens work togetherto initiate a new colony (pleometrosis). However, as the colony begins to grow, workers kill or expel all but one of the queensto make the colony monogynynous. Pleometrosisis believedto rarely lead to permanent polygyny (H6lldoblerand Wilson 1977,Rissing and Pollock 1988). Secondarypolygyny developswhen extra queens are adopted or fused into an establishedcolony that was initiatedby a single queen(haplometrosis). Primary polygyny is usually thoughtto be selectivelyadvantageous because co-operative queens may have a higher survival rate duringthe early foundingperiod; however, the existenceofsecondary polygyne presents a dilemmafor Hamilton's kin selectiontheory (Hamilton 1964 a,b). As mor€ queensare addedinto the colony,there will be a decreasein the overallnesfinate relatedness and in the individual queens'relative fitnessresulting from raising the progenyof adoptedqueens. Why should a resident queen and workers in an establishedcolony acceptadditional queens?Why don't workers on one matriline kill other queensso that their mother will be able to maximize reproductivesuccess? The behavioralinteractions and geneticrelatedness

I Hymenoptera:Formicidae Department of Genetics,Texas A&M University, CollegeStation, TX77g43 3 Departrnentof Entomology, TexasA&M University, College Siation, TX77g43

２７ of co-existing queenswithin a colony mustbe examinedto help explainthe evolutionand maintenanceof polygyrous social organization. During the last two decades,considerable effort hasbeen made to studyrelatedness in polygynous pattems colonies. In someant species,polygynous queens witiiin a colony havebeen reportedto be closelyrelated (craig anocroiiei tszs, Hobr.. 19g6,Douwes et al' 1987' Pamilo 1991). In thesecases, established queens might increasetheir genetic gutpul !y acceptingrelated individuals such as daughters,grand--

MATERIALS ANDMETHODS

Ten monogynousand 24 polygynouscolonies of red imported fire ants were collectedin CollegeStation, Texas. The monogynousqueen, which was readilyidentified by her physogastriccondition and the worker sizefrom her colony (Greenberget al. 1985), 15 workersfrom eachmonog)mous colony, and all the dealates(dealate numbers ranged from 7-56)in polygynecolonies were used to scorethe microsatelliteloci. Twenty-threepolygynous colonieswere collected from six sites in the Brazos Valley Are4 Texas, for queenrelatedness. Only neststhat were separatedfrom other nests by more than 2.Omat any one site were collected to avoid sampling a colony fra ent, as describedby Goodismanand Ross(1997). Colonieswere drip-floatedand hansferredto plastictrays containing two or more 150x 15- petri disheswith a 3-mm-thickcastone floor (Pemaco,St. Iouis, MO). Colonieswere left undisturbedfor the 4 weeksfollowing their collectionto allow dealatesto separateinto groups. Dealateswere removed from the dealategroups within the nest. The head and thorax of each dealatewere removed and stored at -80oCfor later DNA extraction. Abdomensof dealatesfrom colonieswere dissectedto dete ine the mating status of each dealate and to confirm the polygynous condition of the colonies: when the dealateis inserninated,the spe athecais opaqueand milk-white; when the dealateis uninseminated,the spe athecais transparent(Tschinkel 1987). In the text, inseminateddealates are referred to asqueens. GenomicDNA was extractedfrom 1.0 g of S. invicta brood (eggs,larvae, and pupae)from both monogynousand polygynouscolonies individually using equilibrium centrifugationin a CsCl-ethidiumbromide adient(Maniatis et al. 1982). Mbol fra ents of S. invicta genomewere clonedinto the BamHl cut plasmidvector, pUC19,by standud procedures(Maniatis al. 1982). The genomic library was screenedby colony 32P-labelled-et hybridization using synthetic (dA-dC)n oligonucleotides as a probe. Hybridization was conductedovemight at 42"C. Membraneswere washedtwice in a low stringencysolution (lX SSPE, 0.1%SDS) for 15 minutes at room ternperatureand autoradio aphedon an intensifyingscreen ovemight. Positivecolonies were picked with a Pasteurpipette. Well-isolatedcolonies were grown in LB-amp media ovemight and plasmidDNAs were extractedby standardplasmid mini-prep techniques (Maniatis et al. 1982). DNA sequenceofthe insertwas dete ined by Sangerdideoxy sequencing on an ABI Prism 377 DNA sequencer.Insert DNA sequencewas determinedin both directions usingforward and reverse vector PUC/IVfI3 primers. Additionalprimers were synthesized, as needed,for primer walking across large DNA fra ents. Sequenceswere analyzed usingMacVector Version 5.0 software(IBI, New Haven,CT, USA). PCR primers complementaryto unique sequencesflanking eachmicrosatellite locus weredesi ed usingthe programOLIGOv4.O and synthesized by GenosysCompany (Bio- SynthesisInc., P.O. Box 28,Lewisville, TX 75067-0028). TemplateDNAs from monogynousqueens and workers were preparedusing the QtA amp tissue kit (QIAGEN, 28159 Avenue Stanford, Santa Clarita, CA 91355), following the manufacturer'sinstructions. The PCR reactionswere carriedout in a 25-pl reactionmixture containing20mM dNTPs,lX Taq buffer containingl.5mM MgCl2,5pM reverseprimer, 32p-dATp 0.625unit orTaq polymerase,5pM end-labeledforward primer, and 50ng ternplate DNA. Thermal cycling parameterswere standardizedfor all primers and tomplate combinations with specifiJ annealing temperatures:two minutes initial denaturation at 95"c followed by 35 cyclesof 95.d for i5 seconds,4J seconosat the primer specific.armealing tfiiperature and 72oc for 45 seconds,tt" t*t-.v.le was l0 minutes extension at 72c. Aliquots of 5pl were taken from the reactions and electrophoresed on a 6%opolyacrylamide gel, followed by autoradiography.-from- DNA extracted from the queen and 15 wo.i", pupae each of the ten monog)mous colonieswere used to verify the inheritanceof the microsatellitemarkers, and about 500 dealates from polygynous colonies were used to evaluate the level of polymorphism' Microsatellitemarkers that were polymorphicwith four or more alleles wereselected for furtherestimation of geneticrelatedness oipolygyrous queens.Tests for conformityof the observednuclear genotlpe frequenciesto^tttos!-expect"d.rrrdo g*dy- Weinbe_rgequilibrium (HWE) were conductedusing chi-squaretests aithe critical value for cr=0.05. Five hundred and sevenpolygynous queenswere genotypedat four loci, and about were Js-o-salnnfes Senotypedfor the microsatellitevI-Ilocuj-by PCR using radioactive labeledprimers' TemplateDNAs were preparedfrom samplesof ireadand thorax using the pcR QIAamptissue kit, andthe reactionswere performed as described above. Geneticrelatedness and F-statisticswere estimatedusing the softwareprogram RPLATIDI{ESS 4.2c (Goodnightand eueller 1994). colonies-wereweightedequally whertcalculating the relatednessamong the dealatesand queensin the samJcotony. eit available individuals were used to estimatethe averageuilel" fr"quency in the reference population. The standarderrors for estimateswere obtainedbyiack*niirng over colonies (or groups). At t-test was usedto determinewhether the point estimateaiffeied from zero.

RESTILTS.

About 2,000 recombinantclones were screenedusing a (dA-dc)" radiolabeled probe, and 30 positive clones were isolated. Among ttrese, tg microsatelliteswere identified' Potentialprimer pairs for 12 of the microsatelliterepeats were designedusing theMacvector program. Five of the microsatellitesthat were able to be amplifiedin fire ant DNA and were polymorphicwith numbersof allelesfrom four to sevenwere selectedfor furtheranalysis. Primer sequences,annealing temperatures, predicted lengths ofproducts, andallele frequencies of eachlocus are presented in Tablel. Analysis of the expectedand observedlevels of heterozygosity(He, Ho) for each locusamong polygynous queens is given in Table2. The samplesize of N=150was used for analyzing microsatellites M-I and M-II loci and N=517 for the other microsatellites. Expectedheterozygosity for the five polymorphic loci rangedbetween 0.53 and 0.g2. The allele frequenciesfor polygynous queens,monogynous queens and males are also listed in Table 2. The genotlpes of males were inferred from the genobpes of workers in each monogynouscolony. Becauseonly ten monogynousqueens were usedin this study,the sample size is insuffrcient to test for genetic differencesbetween polygynous and monog)mousqueens. Genotypicfrequencies at eachpolymorphic microsatellite locus in polygynous queens fit Hardy-weinberg expectationsat a 5o/olevel, consistentwith Mendelianinheritance of variation in microsatellites(Litte and Luty 1989, Tautz 1989, Weber and May 1989). Progenystudies of ten monogynouscolonies revealed that all worker pupaein a monogynouscolony had a common allele from the presumedfather and anotherallele from their mother in expectedMendelian ratios. Microsatellite genotypesof worker pupae in each monogynous colony indicated that each monogynousqueen was inseminatedby a single male, and that the workers were all offspring of the queenand her mate.

０ R :T C A T rC G C T G A C A A T C G A C A G

M ‐II F :C A A T T A C G T C T C G G T T A T C G A C T C (T G j17 249bp 52oC 4 R :G G G T T G A T C G C A A A A C G A W G

M ―III F :A T T G T A C G A G C G G G A G A A G C A C (A C ), 2llbp 57oC 4 R : T C G A G A T C G A C C A G G T A G G A A C A C

0 M ‐IV F :C G A G T C T C C A C C W G G A A A A G IT G ),7(A G )25 27 8b p 52 C 5 R iC T C A T C C C C A C A A A G T T C A T T G

0 M ‐V F :G T G G A A T G C A C A T C T A A A C 100(A C )(G C ) 3 97bp 56 C 7 R :G G A A A C G C T G A C W T C G T C R ep eat

TABLE2. AlleleFrequencies, Heterozygosity, and chi-square f testfor GenotypeFrequencies t" n*n Utt"r"t"tttrc AlleleFrequencies Heterozygosity Locus A B C D E F G He H O f P

‐ M I P queens N=15o O.3080.043 0.052 0.396 0.201 0`69 0.7 1 3.93 0.70

‐ M II P quecns oNイ90) 0.315 0.038 0.317 0.331 0 169 0 .67 11.80 0 107 M queens (1ギ= 10) 0.100 0.150 0,350 0.400 a M ales cN= 10) 0・400 0.000 0 300 0,300

‐ M III - P qucens oヽ 490) 0 586 0.087 0.343 0.002 0 .55 0 .53 3.9 1 0 .70 M queens G 寺= 10) 0・250 0.100 0.650 0.000 = M aleぎN 10) 0 .100 0 .100 0 .80 0 0 .000 M ―IV P queens cN=490) 0.018 0.643 0.088 0`233 0.018 0.53 0.52 12.56 0.32 M queens N =10) 0.400 0.550 0.000010500.000 M ales4 0N=10) 0.600 0.100 0.200 0.000 0.100 M ‐V = P queens o ヾ490) 0 。106 0 187 0.135 0.272 0 ,07 1 0.22 1 0.009 0.82 0.79 12.02 0 ,30 = M quecns N 10) 0.0 50 0 .250 0 .400 0 ,050 0.150 0.100 a M ales cN= 10) 0 .000 0.600 0.200 0 .000 0 .100 0 ,100 " The genotypeofmale wasinferred from the genotypeofworkers andqueen in a colony

Ten colonies containedseparate dealate groups in which the dealatenumber ranged from 4 to 35; 25% of the dealateson av€ragewere uninseminatedover all samplesstudied. The correlationcoefficient betweenthe numberof matedqueens and tmmateddealates was 0.703(P < 0.001,n = 20). The geneticrelatedness of dealatesand mated queens within a colony were 0.0245+ 0.021 (SE) (samplesize = 490) and 0.0203t 0.0153(SE) (sample size : 350), respectively, not significantlydifferent from zero. Relatednessof dealate femalesand matedqueens within the aggregatedgroups also was not significantlydifferent from zero' The relatednessofqueens bitrveen algregated groups, corielation ooefficients for-relatednessofdealates, and matedqueens betiJen aggregatedgroups :-0'373,P:0'289,n=l0fordealatesandr=-0.0975,p=b.so,i=i0formateoqueens)within a corony(r confirmed that there was no sigrificant relationship between and within the aggregated group dealatesor queens.. Arelativelyhigh estimaieof relatednessin orr. groop did not imply a low estimateof relatednessin another(Table 3).

TABLE 3' Relatedness of Dealatesor Mated Queenswithin a Colony or within and between AggregatedGroups in a Colonv. 里理堅型型ies N o.of lndi宙duals

Dealates 0,0245±0.002 1 23 490 Mated Queens 0.0203±0,0 153 20 350 Dealates Within Groups 0 0043±0 .0367 10 164 0 .0 189±0.0249 10 139 Mated Queens Within Groups 0.0 108±0.042 1 10 120 0.0330±0.0289 10 10 1 Dealates ‐ Between Groups 0.0243±0 .0234 10 164 :139 M ated Between 0,0342±0.0244 10 120 :10 1

All -_ the twenty-threecoronies were collectedfrom the Brazosvalley Area. The overallF,1 was 0.0244t 0.0149(SE) (samplesize: 491),not significantlydifferent from zero at a = 0.05 level, indicating that fire anis in polygyne colonies in the Brazos Valley areacan be treated as a single population. There ii no population structureamong the six researchsites. F;, was0.0774 t 0.0691(sE) (samplesize = qst), andwas not sigruficantly differentfrom zerowhen estimated over all ioci ani from all ttrestuaieo individuals.

里翌里塾匹迎野1■10n cOefflcients betteen N ttber of and R elatedn ess. Relatedness ConelationCoeffi cient(r) p-value No.of comparisons

０ Betweennumber of matedqueens 0 .703 く0 .00 1 and unmatedqueens

４ Betweennumber of dealatesand ‐0 ,136 0 .22 relatednessof dealatesin a colony

令つ Betweennumber of matedqueens ‐0 .0 14 5 0.21 Ｚ々 andrelatedness ofmated queens

０ Relatednessof dealatesbetween two -0 ,373 0.29 aggregatedgroups in a colony

０ Relatednessof matedqueens between ‐0 .097 5 0.86 two aggregatedgroups in a colony

Between relatednessof mated queens 0 .32 0.32 and relatednessofunmated ealates

２ The correlation coefficient between colony queen number and relatednesswasi - 0.136 (P=0.22;n:24'1, not significantlydifferent from zero, indicatingthat relatednessof dealateswithin a colony did not changeas dealatenumber changed(Table 4). Tests for correlationbetween mated queennumber and relatednessof mated queenswithin a colony also showedthat there was no sigrificant correlation between mat€d queen number and relatednessof mated queens(r-0.0145; P=0.21; n=22). The results confirmed that relatednessofdealates or matedqueens did not changeas the numberofdealates or mated queenschanged (Table 4). The correlation coefficient betweenrelatedness of mated queensand relatednessof unmateddealates within a colony was 0.32 (P = 0.32; n: 18). Therewas no significant association between these two values, indicating that a relatively high estimate of relatednessamong mated queensdid not imply a low estimatefor unmateddealates (Table 4).

DISCUSSION

Microsatellites are promising geneticmarkers for S. invicta goretic studies. Rosset al. ( I 999) comparedsix classesof geneticmarkers in describingthe ge'neticstructure of the imported fire ant, S. invicta, and concludedthat microsatelliteshad the greatestproportion of total variance and were the most effective markers for detecting genetic differentiation, Our progeny study of single queeNrcolonies confirmed that microsatellitesin S. invicta are codominantly inherited and selectively neuhal. Analysis of the genotypic frequenciesof eachpolymorphic microsatellite locus indicatedthat the observedallele freque,nciesfit the Hardy-Weinbergexpectation for autosomalalleles. Our earlier studies (Che,nand Vinson 1999) showed that co-existing queensin polygynouscolonies usually aggregatedpeacefully to one or more groups. Workers were obserrredfrequently to pull and push queensback into the aggregationwhen a queenbegan to move away. The studies (Chen and Vinson 2000) also showed that the co-existing queens were mutually tolerant but could exhibit a form of competitiveness. This competitionwas characterizedby the formation of linear attractiveranking to workers. The queen that was most attractive to workers was the queen that had a high frequency of trophallaxis. The hophic advantageresulted in a greater re,productivesuccess of the dominant queen as evidencedby her higher ovipositional rate. Previous studiesposed an interestingquestion about relatednesswithin a queenaggregation and betweenthe queen aggregationsin the samenest. We hypothesizedthat queenswithin the sameaggregation were more related than queens from different ag$egations. This study confirmed the occrurenceofqueen aggregations,and amongthe studiessamples ofpolygynous colonies, ter of 24 colonies (41.7%) yielded more than one dealatefemale group, which rernained stable in undisturbed colonies. Nevertheless,analysis of relatednesswithin aggregated groupsshowed that the aggregationof dealatefemales or queelrsto a group was not based on their relatedness.This study showeda near zero relatednessamong dealate females and queenswithin and between groups, as well ari zero relatednessof queenswithin colonies. The results of this study suggestthat queenaggregation might be a mutualism amongthe nestmatesfor maintaining a stable colony associationand a high colony efficiency. It is unlikely that kin selection plays q major role in mediating the queen aggregationin S. invicta. However, competition amongco-oristing queensfor worker attention could result in workers rearing rmrelated queens,which could promote the evolution of polygynous social organization. Relatednessof dealatefemales or queenswithin colonies in the BrazosValley Arqa was not significantly different from zero. Hence, this indicates that queens moved randomlyto form the permanentpolygynous colonies, and that coloniesfrequently accepted unrelatedqueens from outside colonies. This result is consistentwith previous findings that nestnate queensin introduced polygynous s. inviaa are not closely related and Fletcher1985, @oss Rosset al. 1996,Goodisman and Ro_ss l99s). polygyn; *uy o""* through adgPtionof reproductive queensinto establishedcolonies tri"t orig;:"6 wilh a rew closely related quee,ns(Keller 1995). As more newly mated queens-of aie aoopteo into a colony, quee'n relatednesswill decreasedue to the inEoduction new alleles from non-ne$rnate males. Holever, our results did not show a direct relationship befween either dealate female or queen number and relatedness,as relatednesswars ""ro "ll"n in the smallest colony. studies of the formationof polygynouscolonies in theU.S. suggestedthat most new colonies !olre4r91q may begin as buds of pre-existing polygynous"loionies and Porter 1989). 6vargo Becausethe-pre-existing queens were unrelatio,the queensforming a new bud must also be unrelated. If the newly formedcolonies accepted newly matedqueens from outside colonies,then the near zero relatednessof queenswithin coionieswould be retained. The results of the near zero F,s values of this study indicated that queensin polygynouscolonies in the BrazosValley Area could be considlredas a singlepopulation with high gene flow among colonies. The near zero F, also suggestedthat colonies acceptedqueens from outside colonies. Nonacs(1993) explained-thatthe low genetic relatednessofqueens within socialinsect colonies may be due to the changeofbiological environment' He suggestedthat newly matedqueens choose to remain in tlieir natal nest or enter anothernest, rather,than through independentfounding due to the pressureofhigh population density. Initially, the mated queenshad widef available open habitat, io independenl-foundingof the monogyne w,ur conrmon. As ihe appropriatenesting sites becomefilled, the ability for initiating a new colony is limited. As a result, newly mated queensreentered their natal nestsor searchedfor other nearbynests. Workers initially are glpected to accept only new queensoriginating from their own nest, but the1 nestmate discrimination abilities weakenas queennumber increases@oss et al. 1996). The loss of nesunatediscrimination would presumablycause workers to rear unrelatedqueens thereby causingthe relatednesswithin the colonies to decrease.Nonacs's suggestionmay help to explain the rapid increaseof polygynous forms in Texaspopulation. Ilthough our results andprwious works @ossand Fletcher 1985, Ross et al. 1996,Goodisman and Ross l99g) concludedthat queenswithin the polygynous colonies are not related and that unrelated que€'lrsv/ete frequently accepted into the colonies, direct evidence for genetic factors 9p€rating in polygyre evolution were not provided. The loss of genetic iariation in the introduced S. invicta infened that ge,neticfactors, as well as behavioral and ecological factors, might be involved in S. invicta evolution. Further experimentsin specific lene mappingor relatednessanalysis from different populationswill be helpful.

ACKNOWLEDGMENT

we thankDr. Dawn Gundersen-Rindal(IBUpsI, usDA, Beltsville,Maryland) and Dr'. craig coatef (Departmentof Entomology, Texas A&M Univenity, iexas) for fyiewing an early draft of the manuscript. This work was supportedin part by a legisiative initiative,State of Texas.

LITERATURE CITED

Bourke,A. F. G., andN. R. Franks. 1995, SocialEvolution in Ants. princetonuniversitv Press,Princeton chen' Y. P., and S. B. vinson. 1999. Queensattactiveness to workers in the polygyrous form of the ent solenopsisinvicta (Hymenoptera:Formicidae). Arm. Entomol. Soc. Am. 92:578-586.

４ Chen,Y. P., and S. B. Vinson. 2000. The effectsof quee,nathactiveness to workers on the queennutritional status and egg production in the polygynous Solenopsisinicta. Arm. Entomol.Soc. Am. 93:.295-302. Craig R., and R. H. Crozier. 1979. Relatednessin the polygynous andMymecia pilosula. Evolution33:335-341. Crozier,R. H., and P. Pamilo. 1996. Evolutionof SocialInsect Colonies; Sex Allocation andKin Selection.Oxford University Press, Oxford, UK. DeHeer,C. J., and K. G. Ross. 1997. Lack of detectablenepotism in multiple-queen colonies of the fire ant, Solenopsisinvicta (tl:ymenoptera:Formicidae). Behav. Ecol. Sociobiol.40: 27 -33. Douwes,P., L. Sivusaari,M. Niklasson,and B. Stille. 1987. Relatednessamong queens in polygynousnests of the arr/.Leptothorax ace,lorum. GeneticaTi:23-29. Fletcher,D. J. C. 1983. Threenewly discoveredpolygynous populations ofthe fire ant, Solenopsisinvicta andtheir sigrrificance.J. Ga.Entomol. Soc. 18: 537-543. Glancey,B. M., C. H. Craig,Y. P. M. Stringer,and C. E. Bishop. 1973. Multiple fertile queensin coloniesof the importedfire ant,Solenopsis invicta, J. Ga.Entomol. Soc, 8:237-238. Glancey,B. M., J. C. E. Nickerson,D. Wojcik, J. Trager,W. A. Banks,and C. T. Adams. 1987. The increasing incidenceofthe polgyny form ofthe red imported fire ant, Solenopsisinvicta (Ilynenoptera: Formicidae),in Florida. Fla. Entomol. 70: 400- 402. Glancey,B. M., and C. S. Lofgren. 1998, Adoptionof newly matedqueens: A mechanism for proliferation and perpetuationofpolygynous red imported fire ants, Solenopsis invictaBuren. Fla. Entomol.71 : 581-587. Goodisman,M. D., and K. G. Ross. 1997. Relationshipof queennumber and queen relatednessin multiple-queen colonies of the fire ant, Solenopsisinvicta. Ecol. Entomol.22: 150-157. Goodisman"M. D., and K. G. Ross. 1998. A test of queen recruitnent models using nuclear and mitochondrial markersin the fire wfi, Solenopsisinvicta. Evolution 52: A16-142. Goodnight,K. F., andD. C. Queller. 1994. Relatedness4.2C. GoodnightSoftware. Greenberg,L., D. J. Fletcher,and S. B. Vinson. 1985. Differencein worker size and mound distribution in monogynous and polygynous colonies of the fire ant SolenopsisinvictaBluren. J. Kan. Entomol.Soc. 58: 9-18. Hamilton,W. D. 1964a. The geneticevolution of socialbehaviour I. J. Theor.Bill. 7: l- 16. Hamilton, W. D. 1964b. The geneticevolution of social behaviow IL J. Theor. Biol. 7: t7-52. Herbers,J. M. 1986. Nest site limitation and facultativepolygyne in the ant Leptothorar longispinosus.Behav. Ecol, Sociobiol.19 : 115-122. Herbers,J. M. 1993. Ecologydeterminants of queennumbers in ants. .lz: QueenNumber and Socialityin Insects(ed. By L. Keller). pp.262-293. Oxford UniversityPress, Oxford. H6lldobler, B., and E. O. Wilson. 1977. T\e number of queens:an important trait in ant evolution. Naturwissenschaften.64: 8-15. H6lldobler,8., and E. O. Wilson. 1990. The Ants, Belknapof HaryardUniversity Press, Cambridge. Hung, A. C. F., S. B. Vinson, and J. W. Summerlin. 1974. Male sterility in the red impodedfire ant,Solenopsis invicta. Ann. Entomol.Soc. Am. 67:909-912. Kaufrnann,B., J, J. Boomsma,L. Passera,and K. N. Peterson, 1992. Relatedn€ssand inbreeding in a French population of the unicolonial ant Iidomyrmex humilis (Mayr). Ins. Soc.39:195-213. Keller, L' 1995. Social life: The paxadoxof multiple-queencolonies. TrendsEcol. Evol. 10:355-360. Krieger,M' J' B., and L. Keller' 1997. Polymorphismat dinucleotidemicrosatellite loci in fire ant (Solenopsisinvicta)population. -A Mol. Ecol.6:997_999. Litte, M', J, Luty. g31t -. 198.9. hlpervariable microsatelliterevealed by in inho amplificationof a dinucleotide.Am. J. Hum. Gerct.44i 3g74}l. Maniatis, T.' E. F. Fritsch, and J. Sambrook. r9.gz. Molecular cloning, a laboratory press, .^T9. Cold SpringHarbor Cold SpringHarbor, New york. Mirenda,J. T., and S. B._vinson. r9g2. Singleand multiple queencolonies of imported fire antin Texas.Southwest. Entomol.l: 135_141. Monill, w' L. production 1974. and flight of alate red imported fire ant. Environ. Entomol.3:265-271. Nonacs,P' 1993. The effects of polygyny and colony life history on optimal sex pp. I 10-130. L. (ed.) Tvlsment' ^/n: Keier eueenNumber and Socialityin Insect. Oxford University press,Oxford. Pamilo, P. 1991' Evolution of colony characteristicsin social insects. II. Numberof reproductiveindividuals. Am. Nat. l3g: 412_433. Porter,s. D., A' Bhatkar,R. Mulder, S. B, vinson, and D. clair. 1991. Distributionand density of polygynous fire ants (Hymenoptera:Formicidae) in Texas. J. Econ. Entomol.84:866-874. Rissing, pleometrosis s. w., and G. B. Pollock. r98g. and polygyny in Ants. wesMew, Boulder. Ross,K. G. 1993. The breedingsystem of the fire ant, sorenopsisinvicta: Effectson colonygenetic structure. Am. Nat. l4l:554_576. Ross,K. G., andD. J. c. Fletcher, 1985. comparativestudy of geneticand social structure in two forms of the fire art Solenopsisinvicta Qlymerrof,tera:Fomricidae). Behav. Ecol. Sociobiol.l7 : 349-356. Ross,K. G., and L. Keller- 1995. Ecologyand evolutionof socialorganization: Insights from llre ant andother highly eusocialinsects. Ann. Re. Ecol. Sfst. 26: 631-656. Ross, K. G., M. J. B. Krieger, D. D. Shoemaker,E. L. Vargo, *i L. K"ll.r. lgg7. Hierarchicalanalysis of geneticstructwe in nativefire anlpopulations:Results from threeclasses of molecularmarkers. Gene. Soc. Am. lg97:6q_655. Ross,K, G., D. D. Shoemaker,M. J. B. Krieger, C. J. DeHeer,and L. Keller. 1999. Assessing ge'neticstructure with multiple classesof molecular markers: A case study involving the introducedfire ant solenopsis invicta. Mol. Bio. Evol. 16: 525- 543. Ross,K. G., E. L. vargo, andL, Keller. 1996. Socialevolution in a new environment:The casedf introducedfire ants. proc.Natl. Acad.sci. USA 93:302r-3025. Tautz,D.- 1989. Hypervariabilityof simplesequences as a generalsource for polymorphic DNA markers.Nuc. Acids Res.lZ: 6463-6471. Tschinkel, w. R. 1987. Fire ant queenlongevity and age: Estimation by spermdepletion. Ann. Entomol.Soc. Am. 8O:263-266. Vargo, E. L., and S. D. Porter. 1989. Colony reproductionby buddingin the polygyne form of the fire ant, Solenopsisinvicta. Ann. Entom.Soc. Am. g2;307_313. weber, J., andP. May. 1989. Abundantclass of humanDNA polymorphismwhich canbe tlped usingthe polymerasechain reaction. Am. J. Hum. Genet.44: 3gg-396.

６ V O L .2 8 N O .1 S O U T H ヽ電S T E R N E N T O M O L O G IS T ヽLAぷこ2003

DECOMPOSITIONSTUDIES, WITH A CATALOG AND DESCRIPTIONSOF FORENSICALLYIMPORTANT BLOW FLIES (DIPTERA:CALLIPHORIDAE) IN CENTRAL TEXAS ", b, Felix M. Tenorio JimmyK. Olson CraigJ. Coates" TexasA&M University,Department of Entomology

ABSTRACT

This study was conductedin Brazosand BurlesonCounties of Central Texasusing 20-25 lb swine carcassesas bait. The pig carcasseswere allowed to decomposeto skeletonization,and the blow flies and other arthropodsassociating with the carcassesat the variousstages ofdecomposition were collectedand identifiedin orderto catalogtheir occurrence. Twenty-threefamilies of insectsand mites were identified.The speciesof blow flies associatingwith the carcasseswere studied closely in orderto makepreliminary predictionsabout the seasonsin which they occurin this areaofTexas.

INTRODUCTION

Forensic entomology makesuse of iniects' predictablelife cycles to help provide clues in deathinvestigations. Insects that feed on carrion are so specializedthat they occur only when environmentaland biochemical conditions are perfect. The mere presenceof somespecies can provide detailed information about the location,time, andconditions of a corpseor carcasswhen the collectionwas made. Blow flies (Diptera:Calliphoridae)are the insects of most importance to medicolegal entomology becausethey are usually the first to colonize a carcass,often within minutes of exposure(catts 1992, Greenbergl99l). The speciesof blow flies present in a given area can be used in a court of law to help establish such things as geographiclocation, time of year, andpostmortem interval @MI). Presently,Hall's Blow Flies of North America (Hall 194g) and the chapteron calliphoridae in A Manual of Nearctic Diptera (McAlpine 1993) are among the best souraesof information about the distributionof many forensicallyimporlant blow fly speciesin the u.S, However,neither of thesesources is specificto central rexas. The importance.of catalogingthe most commoncarrion-associating species and succession patternsofthese speciesin CentralTexas was stressedmost recently by Byrd and Castner (2001).They statethat the biogeoclimaticzone in which a set of remaini are found will have a major impact on the types and speciesof insestspresent as well as the seasonal availabilityof thosespecies. The project describedherein was intendedto provide informationabout the blow fly specie-spresent in Brazos and BurlesonCounties of Cenhal Texas during different seasonsofthe year,and to catalogthe speciesofcanion-associating arthropods in the area. 軽穏禄鞘脳盤盟穏紺誂 繊 MATERTALS AND METHODS

In an effort to representactual death scene settings, two sites were chqsen to conductthis srudy.The first site was on trrewest campusii r.i., aali'university in college station (Brazos co.), Texas, adjaceni to it, TAMU rraffiito Research Laboratory'The site was,a grls.sy areawittr only one tree in the immediatevicinity and otherwisesurrounded and partiallyshaded by uuitoings.This site was chosenas the urban site for the study. The second site was iocated approximatelyz.s tilomete.s south, southwestof Snook(Burleson co.), Texas,and representeda more rural settingcommon to central Texas.This site was in a woodei *.u undshaded for "r"ri"itrr.'0"y. The two siteswere_approximately 25 kilometersapart. wild adult and larval calriphorid fly populationswere surveyedusing swine carcassesas bait' carcasseswere ptovided bathi iexas earur uniuersi{iledical school under TAMU Laboratory Animil use protocorsallowing io. iirt*-iiuring between researchprojects. The pigs were freshlykilled andnot chill;d prio. to U"ingset out in the Iatemorning or earlyafternoon (between hours 10:00and 14:00r,ou..l. euin carcasswas coveredwith a wire cage prevent to. predationby vertebrate,.uu.ng..r'*J wasmonitored at the siteover a period of time to allow for completedecompositioi t"rt"Llr remains. The possibility that some of the arthropodscollected were the ,rruit or previous carcassesat that particular site was taken into ionsideration, *d * ;.;;was made to this from lt"Y"nt happeningby allowing eachcarcass to reachthe point ofikeletonization. Each time, the area was allowed to resi for at least 5-z days u"ror" *otrr"i carcasswas placedin the samelocation. Adult flies were collectedby using a short-handtedsweep net, a portableinverted conefly trap,sticky traps, and.yellgyp*.gugr. Fliescollected *ing tt " sw-eepnet andthe cone trap were killed, pinned and identified.Fli'es collected on ihe sticky traps and in yellow pantraps were counted and identified. The invertedcone trap, sticky traps,and yellow pan trapswere passive methods of collectionand used mainly to trap any flies andott.t a.ihropod,*rut ,o. not caughtwith the sweep.net.Sticky traps were usedwhen fly activity *u, ih" gr."t"rt. Tlr" .one trap was baitedwith rotten beefheart, and the yellow pan trapswere setout next to and aroundthe carcassesovernight and filled with soapywater to trapthe insects. Fly eggs,larvae, and pupaewere collectedon dayswhen they were notedto be presentusing a.microspatula or forceps.The microspatulawas used to scoopup samplesof eggmasses andfirst- andsecond-instar larvae. The forcepswere used to coilectthird-instar larvaeand pupae. Only - smallrepresentative samples of eachlife stagewere taken from the carcasses. Thesesamples were immediatelyplaced on rearingmedia, Friskies Fancy Feast@ Sliced chicken Hearts and Liver Feastfr moist cat food (Friskies petcare company Inc., Glendale, cA 91203)and kept at 22'c (+loc) in BioeuiR@(Gardena, cA 90248-3602) mosquitobreeder containers filled with 1.5 inchesof vermiculiteto facilitatepupation. Samples of larvaewere treated in the samemanner with approximatelyhalf oieach sample killed in pupae, near boiling water. harvestedafter all larvaemigraieJ fi"; the carcass, weretreated in the samemanner. _ Each specimenwas identified using morphological characteristicswhen the characterswere viewable. Third-instar larvae were riared-in the laboratory and identified using taxonomic \gvs br.stojanovich et al. (1962) and Debangana creenuerg(19g9). Adnlts were identifiedusing- thc just .samekeys noted as well is ttrore in ) uoirot oy NearcticDiptera (McAlpine 1993). Samplesofthe larvaeand adultsofother arthropodspresent on the carcasseseach day werecollected and pinned or preservedin 80% ethanol.These samples were identified to family and genususing lz Introduction to the study of Insects@orror et ar. 199?),How to Know the Immature Insects (chu and cutkomp 1992), and A Manual of Acarologt (Kmntz 1978). SuccessionPatterns ofthe arthropodspresent for each carcassstudy are basedon written observations,sticky board collections, collections of larvae and eggs, sw€ep net collections,pitfall collections,yellow pan collections,and the rearingofeggs, larvaeand pupaeto adulthood.The patternswere based on the presenceor absenceofarthropods on eachday of observation.Collections were not madeon dayswhen the temperaturewas too cold to allow arthropodsto be presentand on daysofheavy rainfall.

RESULTSAND DISCUSSION

Occarrenceof Calliphondae in Brazos and Burleson Counties:Table I showsthe speciesofblow flies collected at each study site, and takes into accountboth larval and adult stagesof eachspecies. Most of the Calliphoridaespecies collected were presentin both settings.Phaenicia cuprina Wiedemannwas not collectedat the rural studysite. This specieshas beendocumented to prefer carrion locatednear humandwellings (Byrd and Castner2001 ). Table2 showsthe blow fly speciescollected by season.

TABLE l. List of CalliphoridaeTaxa Collected from SwineCarcasses from October1999- october 2000 at Two study Sitesin Brazosand Burlesoncounties of central rexas. Occurrenceat study sites Genus Species Urban Rural 腕 Ｘ Ｘ Calliphora す 吻 Ｘ Ｘ

Ｘ Ｘ Cynomyopsis(= Cynomya) cadaverina

Ｘ Phaenicia cuprtna Ｘ Ｘ eximia Ｘ Ｘ coentleiviridis

Ｘ Ｘ Cochliomyia macellaia

Ｘ Ｘ Chrysomya ru/ifacies Ｘ megacephala

Ｘ Phormia regina

The following distinguishing characteristicswere noted from the different species o{b19y flies occurringin Brazosand BurlesonCounties based primarily on observations of wild populations. cochliomyia macellaria (Fabricius). Representativesof this species typically displayeda dark greenio blue metallic abdomenand thorax with three longitudinai stipei on the dorsum of thorax. In Brazos and Burleson Counties,this speciesoccuned in late spring, throughoutthe summer,and into late fall (Table 2). Memlers of this speciesare

９ usuallythe fint to ovipositon a carcass,and the larvae-formlarge feeding masses. During the summer months, larvae of this species are often driven out from the carcass prematurelyby thoseof C. ruJifacies. Phoryia ^ . regina (Meigen).Adults of this speciescollected in Brazosand Burleson Countiestypigally displayedmetallic dark olive green to metallic dark blue coloration of the thorax and abdomenwith the anterior spiracle-appearingdark orangeor red, and a row of bristlespresent on the radialstem of wing. Durdi the ciurse or thii studv,this species occurredin late fall throughoutlate spring(iable Z). fhis speciesis well establishedin the areaand occurredin greatnumbers in both rural andwban settings.Adults ciustertogether to lay largenumbers of eggsand create huge masses of larvae. chrysomya rufifacies(Macquart). This speciesoccured during rate spring,earry summer, and into the fall months of the year (Table 2). Adults Jf this ipecies arl characterized by a row ofbristles presenton the radial stemofthe wing, u u.igtt *tit" anterior spiracle,white buccalarea, white, pubescentsquamal lobes, and bright metallic greenabdomen and thorax.The anteriordorsal portion oithe tho.a* hasthree small black lines that do not stetch pastthe middle of the thoraxand areoften difficult to seewithout closeinspection. This specieswill anive first at a carcass,but doesnot tend to oviposit firsJ. Femalestend to lay eggsin masseson theunderside ofcarcass (under jaw, underribs, and any otherpart touchingthe ground).Larvae then migrate up*u.d, intoihe carcassand eventuallydrive out populationsof other speciesas notid by other authors(Tantawi and Gre^enberg1993). This speciesis well establishedin CentraiTexas where this studywas performed, is highly competitive,and the larvae form rolling masses.This species competeswith Phaenicia coeruleiviridis (Macquart),c. macellarii, and,p, regina. . chrysomya megacephala(Fabricius). During the course of this stud], only three specimensof this specieswere collected.This speciesdoes not have weil-established populationsin either of the fwo collection sites.Tlie collectionsmade of this specieswere at first accidentaland afterwardsonly madeby close observation.Random sighiings of this specieshave beenmade in rural areasof college Station,Texas (Highwai 39, personal observation).This species was collected in November $able z), urrd udrrlt, ur. characterizedby a darkbrown to black anteriorspiracle and a row ofbristlespresent on the radial stemof the wing. Males of this specieshave largeeyes which take up most of the headarea, while femalestend to appearmore rike c. ruffacies with the eiception of a yellow buccalarea, and dark brown to blackanterior spiracle. Phaeniciacoeruleiviridis (Macquart), Members of this speciesoccur in the late springand early months of the summer(Table 2). Adults form swannsabove a carcassand lay.large numbersof eggs.The larvae of this speciesform huge massesand are very similar in appearanceto P. cuprina andp. eximia (Robineau-Desvoidy).Adults of thii speci-99have a very bright metallic greenthorax and abdomenand sometimesappear bright metallicblue. Males of this speciescan be distinguishedfrom P. eximiaby the presencJof onepair ofreclinatefrontoorbital bristles. Phaeniciaeximia (Robineau-Desvoidy).This specieshas often beenmisidentified becauseof its similarity to P. coeruleiviridis(Byrd and castner 2001). In Brazosand B.f49to" Counties, it replacesP. coeruleiviridis dwing ttie late summer and early fall (Table2). The larvaeof this speciesare almostidentical to P. coeruleiyindrs,References to P. eximia in any of the larval keys that were used in this sfudy were not found. This speciesoccurred in small numbersin both settings.The best characterfor identification found during this study was that males of P. eximia do not have a pair of frontoorbital reclinate bristles. Femalesof this speciesare diflicult to identi$. Ttri best characterfor identificationof this species,at the moment,'isthe seasonwhen it occurs.

４０ TABLE 2. Occurrenceof Blow Fliesby Seasonat SwineCarcasses Located in Brazosand BurlesonCounties ofCenhal Texasfrom October1999-October 2000. Seasonsofoccurrence Winter Spring Summer Fall Genus Species (Dec.-Feb.) (Mar.-May) (June-Aug.) (Sept.-Nov.)

Calliphora livida rcoq)ofixxxx)oqxxxx yicina W xxxx)o(xxx

Cynomyopsis (=Cynomya) cadaverina ffi X)OfiXXXXX

Phaenicia cuprina eximia xxxxxxxxx coeruleiviridis xxxxxxxx

Cochliomyia macellaria ffiffi

Chrysomya rufifacies megacephala xxxxxxxxx

Phormia regina xxxxxxxxxxxxxxxxx xxxxxxxxx

Phaenicia caprina (Wiedemarur).This Speciestypically displays a bronze to metallic light green abdomenand thorax, with stout bristles over the entire body. This speciestends to lay small numbersof eggs,and the larvae are often not collected. Larvae of this speciesare similar to those of Phaenicia spp. but difler in the number of branches on the anterior thoracic spiracles.This speciesoccurs in late spring into the summer,and early fall in Central Texas(Table 2). Cynornyopsis(=Cynomya) cadaverina Townsend.This speciesof bluebottle fly occurredfrom mid-winter to early spring during the courseof this study (Table 2). Adults of this speciesare very similar in appearanc€to C. vicina and,C. livida, having a dull gray to black thoraxand a metallicblue or greenand occasionally pollinose abdomen. Members oftlris genusare distinguishablefrom the Calliphoragenus by the occurrenceofone pair of preintra-alarbristles on the thorax.Females of this speciestended to lay eggsin small clustersor scatteredsingly. Adults arestrong flyers emittinga loud buzzingsound. Calliphora vicina (Robineau-Desvoidy).This speciesof blow fly is highly endophilic.Many specimensof this specieswere caught inside the TAMU Mosquito ResearchLaboratory. Adults of this speciesare robust and strong fliers. Ttie greatest numbersof individuals were caughtin early to mid-spring. Calliphora livida }JalI. The morphology of adults of this speciesis very similar to C. vicina. Specimenscollected in Brazos and Budeson Counties were smaller than C. vicina and C. cadaverina.The abdomenof this specieshas a more violet coloration than that of the other bluebottle species. Larvae of this speciestended to be found in mixed sar.nplesof larvaebelonging to otherspecies. While not asabundant as C. cadaverinaor C. vicina,this speciesis well establishedin the area,This speciesoccurs in the centralTexas arBa(Brazos and Burleson Counties)during the late winter and early spring.

４ ∽就 暫概済と棚 ざ °mpiLd ttm pttous me面oned 樺 縮), け 網 患縄 鞘 覗盤

w い o ut Bn ra l t a la r b rt tle w iut antrB _a 崎「b 、 tte

。ン砲"拘匹/t 協 す 2 … ～ e p p

' Ba!l@3ta light otorcd ' B*lgoata dark @tor€d (reddleh q golden) .HU.dbo€bt (Bbck) ck 'dl bu€ d*oa ' ' btek 2 pd intE€lrr brbtL. 3 p6t tab4tar b.l.d- I I I 1 CeNpM*tna @l>@tad.

F IG .l Specics tree diagram fOr (ユcαttαw /izα,a ″▼↓aa,and てュソた,″α.

妨 競 岳捕 鷲 替 !縛 群 鵠 倉盤 t燃 群 鏑 e盟 綸 & 汗歓 溢 景

dom inant early ar iver during the w inter m onths. O n en c cαttαソer,■α w as the flrst to aF iVe us , ually on the i rst day,and the fem ales iIIIm ediately began to lay eggs.

４２ T A B L E 3. L ist of A 11 0 ther T axa C ollected at B oth C arrion Sm dy Sites in C enttal.T exas O ctober 1999‐O ctober 2000.

C ollect on site Class O rd er F anlilv Genusand Soecies' R ural U rban Ｘ Ｘ Hexapoda つ ″陶 ″ Sarcophagidae Ｘ Ｘ Muscidae (llinnaeus) Musca domes tica Ｘ Ｘ Fannia Ｘ Ｘ Hydrotaea/Ophyra Ｘ Ｘ (Linneaus) Piophilidae P iophila casei Ｘ Ｘ Sepsidae Ｘ Ｘ Phoridae Ｘ Ｘ (Linneaus) Stratiomyidae H ermetia il lucens Ｘ Asilidae Ｘ Ｘ 命′ r∽ Silphidae 々p Ｘ Ｘ Staphylinidae Creophilus maxillosis (Linneaus) Ｘ Ｘ sp.l Ｘ Ｘ sp.2 Ｘ Ｘ Histeridae Saprinus p ennsy lvanicus Paykull Ｘ Ｘ sp. I Ｘ Ｘ Scarabeidae Ｘ Ｘ Trogidae Ｘ Geotrupidae Ｘ Ｘ Dermestidae D ermes t es caninus German Ｘ Ｘ Nitidulidae ita coloz (Linneaus) Omos Ｘ Ｘ Cleridae Necr obia rutipes @ eGeer) Ｘ Tenebrionidae Ｘ Hymenoptera Yespidae Ｘ Ｘ Formicidae Ｘ Chalcididae Ｘ Hemiptera Reduviidae

A rachn ida Acari Ｘ Ｘ Mesostigmata Macrochelidae Ｘ Astigmata Acaridae Crustacea Isopoda ' Many ofthesearthropod speimens werc only identifiedto family

Muscoid flies of the genercOphyra/Hydrotaea, Fannia, and Musca did not occur until after the blow flies were well established,and the larvae of these generawero not found until final collections were made for each carcass. Adult cheese skippers (Diptera:Piophilidae) were presentin large numbersduring the fall months of this study very early in the decompositionprocess, but their larvaewere only found during final carcasscollections. Soldier flies (Diptera:Stratiomyidae)showed the samebehavior as the cheeseskippers, but their larvae, when present,were more numerous.Flies of the family Sepsidaewere collected from a carcassat each study site, but only adults were collected and never in high numbers. These flies are scavengersand are very conrmon on dung (Byrd and castner 2001). In thesecases, they may have been attractedto the carrion, but did not seemto find it a suitablefood source. The speciesofColeoptera associating with eachcarcass were only presentafter the larvae of the blow flies and flesh flies becamewell established.Creophilus maxillosis

４ (Linneaus) (Coleoptera:Staphytinidae) was the dominant speciesofrove beetlepresent at both locationsduring the sunmer months,with otherspeciei of smallerrove beeties.taking over when the temperaturesbecame cooler in the fali and winter. In each case,carrion beetles (coleoptera:silphidae), and clown beetles (coleoptera:Histerida; were b.;; 9:fu the skin (coleoptera:Dermestidae), checkered (coteoptera:tteridai), sap (coleoptera:Nitidulidae),and hide (coleopera:Trogidae)beetles. The only speciesoi checkeredbeetles present at both locationswas Necrobia rufipes (DeGeer),and omosita colon (Liwteaus)was the only speciesof sapbeetles collectid. lti the aduit beetleswere conspicuousand easy to collect, They often wanderedon the carcassand surrounding vegetationduring the daylighthours, especially when the temperatureswere warm. Hide beetles,and otherbeetles of the family Scarabeidaewere rarely collected.When collected, they were found in and under the carcassduring late decompositionand skeletonization. Beetle larvaewere not presentuntil after skeletonizationexcept for one carcass(Rural Setting:1118100-2122/00) that was coveredwith carrionbeetle larvae on Day 20. when the larvaeofbeetles were present, rove beetlelarvae were the mostabundant. They wereoften foundinside tle carcassand in surroundingvegetation. As for other arthropods collected during the study, predatory mites (Acari: Macrochelidae) were more prevalent in the rural setting, and were often seenattached to the skin beetles.They were conspicuousand easily collectedbut did not arrive until the later stagesof decomposition(i.e., puhefactionand skeletonization).A family of stored product mites (Acari:Acaridae)was also collectedin huge numberson severalof the carcassesduring the same stagesof decompositionand were collected as adults and hypopialstages. In summary,this study has helped to show that the speciesof carrion-associating insectsin central rexas vary with seasonand locale. Some of the blow fly species collectedoccurred only at the rural site of colleciion,while otherswere only collectedat the urban collection site. This indicatesthat some blow fly speciesmay prefer one environmentto the otheras previouslynoted in otherstudies (Wells and Greenberg1994, Smith 1986).In addition,some of the speciesof blow flies collectedhavenotpreviously beenrecorded as occurring in central Texas.For example,C. megacephalais a new arrival to Texas.The collectionsof two malesand one femaleduring the courseof this studyare new recordsfor the area. As previouslynoted by Wells and Greenberg(1994), fire ants are an important factor in the calculation of a post-morteminterval. Wheneverfire ants were presentnear the study sites,their foraging area included the carcassesput out for this study. They may delayblow fly ovipositionas much as two dayswhen they arepresent, further supporting the observationsby otherauthors (Wells andGreenberg 1994). Insectsuccession in the areathat was studiedfollowed a setpattern. The flies were alwaysthe first wave,followed by wavesof beetles,more flies, andthen more beetles. The blow flies werenot alwaysthe first insectsto colonizethe carcass.Often, flesh flies arrived first, especiallyduring the fall monthsof the year. Insect activity did not ceaseat any point during the courseof this study, The diversityofspecies was greatestduring the warm months,but evenon the coldestdays of the year,the carcassesdid not fail to be infestedwith sometypes of insects.Decomposition wasretarded by the cold weather,but flies andsome beetles were present. P. caseiwas one of the dominantflies duringthe fall andwinter, when the weatherwas the coldest,

LITERATTIRE CITED

Borror, D. J., C. A.Triplehom,and N, F. Johnson.1992. An introductionto the study of insects.Harcourt Brace College Publishers: Fort Worth,TX. Byrd, J. H., and J. L. cashrer.2001. Forensicentomology: The utility of arthropodsin legalinvestigations. CRC Press: Boca Raton, FL. Catts, E. P. 1992.Problems in estimatingthe postrnorteminterval in death investigations. J.Agric. Entomol, 9:245-55. Chu,H. F., andL. K. Cutkomp.1992. How to know the immatureinsects. Wm. C. Brown Communications,Inc, : Dubuque,[A. Debang,L., andB. Creenberg.1989. Immature stages of someflies of forensicimportance. Ann.Entomol. Soc. Am. 82:80-93. Greenberg,B. 1991.Flies as forensic indicaton. J. Med. Entomol.28:565-77. Hall, D. G. 1948.The blowflies of North America,Vol. IV. A ThomasSay Foundation Publ.,USDA Bureauof Entomol.and Plant Quarantine. Monumental printing Co.: Baltimore,MD. Krantz, G. W. 1978.A manualof acarology,2nd Ed. OregonState Univ. Book Stores, Inc.: Corvallis,OR. McAlpine, J, 1993.Manual of nearcticDiptera, Vol. 2. CanadaCommunication Group Publishing:Ottawa, Ontario. Shewell,G. E. 1993.Calliphoridae, pp. l133-1145.In:J.F. McAlpine(ed.) 1993. Manual of nearcticDiptera, Vol. 2, CanadaCommunication Group Publishing:Ottawa, Ontario. Smith, K, G. V. 1986.A manualof forensicentomology. Comstock Publishing Assoc. CornellUniv. Press:Ithaca, NY. Stojanovich,C. J., H. D. Pratt, and E. E, Bennington,1962. Fly larvae:Key to some speciesof public healthimportance, pp. 125-133.In: Anon. (ed.) 1969.pictorial Keys to arthropods,reptiles, birds and mammalsofpublic health significance.U. S. PublicHlth. Serv.Publ. No. 1955.U. S. Gov.Printing Oflice: Washington,D.C. Tantawi,T. I., andB. Greenberg.1993. The effeci of killing andpreservative solutions on estimatesofmaggot agein forensiccases. J. ForensicSci. 38: 702-707. Wells, J. D. and B. Greenberg. 1994. Effect of the red imported fire ant (Hymenoptera: Formicidae)and carcasstype on the daily occurence of postfeedingcarrion-fly larvae(Diptera: Calliphoridae, Sarcophagidae). J. Med. Entomol.3l: l7l-174.

４ V O L .2 8 N O .1 S O U T H W E S T E R N E N T O M O L O G IS T は 2 00 3

COMPARISON OF SUSCEPTIBILITY OF GEOCONSPUNCTIPESI AND ZTGUS LNEOIANSA TO INSECTICIDESFOR CONTROL OF THE TARNISTIED PLANT BUG

P. Glpn Tillman, JosephE. Mulroonet', and GordonL. Snodgrassa,s

usDA, ARS, crop Protectionand Management Laboratory, p.o. Box 74g rifton, GA 31793

ABSTRACT

Compmisonofthe susceptibility of Geocorispunctipes (Say)and the tamished plant bug (TPB),Lyguslineolaris (PalisotdeBeauvois) to selectedinsecticideswas determinedintopicai, tarsalcontact, and field studies.In both topical andtarsal contact studies, Z. lineolaris wasmore susceptibletoimidaclopridandoxamylresiduesthanG. punctlpes. However, oxamyl wasmore toxic to the pest than imidacloprid. Both insect speciei respondedvery similarlyto fipronil, acepbate,dicrotophos, and lambda-cyhalothrin, all of which werevery toxic to theseinsects. In our field study,lambda-cyhalothrin had an equally negativeimpact on populationsofTpB and G.punctipes concurringwithpreviouslypublished field studies.Results from our laboratoryand other field studiesindicate that oxamyl and imidacloprid would be effective againstTpB while conservingpopulations of G.punctipes for biologicalcontrol of lepidopteranlarvae in cotton.

INTRODUCTION

Miridae) is often a serious tissue-suckingpest of cotton, Gosqypiumhirsutum L. causing abnormalplant growth, f it damage,deray in fruiting, and delayedboll maturity (Hannyet al. 1977). The big-eyedb'g Geocorispunctipes (Say)lHemiptera: Lygaeidae),is a predatorof pest many speciesincluding I/ uliescens andHelicoverpa zea (Boddie) eggsand small larvae (Lingren et al' 1968).Geocoris punctipes alsofeedson plants,increasing ttJfkelihood oftheir survivalduring the absenceofinvertebrate hosts (Eubanks and Denno-l999). Since the TPB and G. punctipes can occur in coffon fields concurrently,selectivity of insecticides with respectto thesetwo insect speciesis an importantissue in anintegratedpest managementprogram becauseinsecticides recommended for L. lineolaris cont ol -uy be harmful to G.punctipes. TPB and G.p unctipes canbeaffected by insecticidesthrough various routes of administration: topical contact on their bodies, tarsal contact with residues of insecticides,'and feeding on insecticidetreated plants and prey (for predator). The organophosphate,acephate, recommended for usein cottonfor TPB controlin 1977,provides effective TPB suppression(Bannister et al. I 995, Reed et al.l997, Robbins et al. l 99g). The pyretbroid insecticideswere first registeredfor usein cotton in 1978,and lambda-cyhalothrin wasused veryeffectivelyagainstTpB (Grahamand Gaylor l9gg, Leonardet al. l9g7), Another organophosphate insecticide,dicrotophos (Burris et al. i 9g6,Graham and Gaylor l9gg, 6onard -^Honiptem: Lypeidae 'Hemipt€n: Miridae 'USDA, ForesrService,20l 'U-SDA, LincolnGrem, Starkville, MS 39759 Agrioltural Research p.O. -Mentlon Service, Box 346, Stoneville,MS 3g776 ofa proprietarypmduct dos not constitutean endorsetnentor reconrnendationfor its useby USDA.

４ et al. 1987,Langston and Schuster 1989), and the carbamate, oxamyl (Micinski 1983)also were usedsuccessfully for TPB contol during this time. Pyrethroidresistance was first detectedin 1993in a field populationof TPB in theMississippi Delta, andresistant insects also weri found to havemultiple resistanceto someorganophosphate and carbamate insecticides (Snodgrass and Elzen1995). Level ofcontrol with organophosphate,carbamate, and pyrethroid insecticides has decreasedover the past severalyears in someareas ofcotton productionwhere resistance to all theseinsecticides has been reported. However, dicrotophos, oxamyl, and lambda-cyhalothrin, arestill usedeffectively against susceptible TPB populations(Bannister et al. 1995,Pankey et al.1996, Reedet al.1997, Robbinset al. 1998,Russell et al. 1998).The new insecticides imidacloprid,an imidazolidinimine,and fipronil, a phenylpyrazole,have shown excellent activity on TPB even against TPB populationsthat are highly resistant to other classesof insecticides(Burris et al. I 994,Bannister et al. 1995,Scott and Snodgrass I 996, Shaw and yang 1996,Teague and Tugwell 1996). Most of the reportedinsecticide research on G. punctipp-sconcentrates on residual toxicity of insecticideswhich has been low for someinsecticides such as spinosad and oxamyl andhigh for otherinsecticides such as malathion and fipronil (Boyd andBoethel 1998, Elzen et al. 1998,Elzen and Elzen 1999,Tillman and Mulrooney2001). The few topical toxicity studiesconducted have demonsfrated that malathionand other organophosphates, fipronil, and cyfluthrinare highly toxicto G.punctipes (Lingren and Rid gway1967 ,Tillman andMulrooney 2001).The singlestudy reported on effectoffeeding on dried residuesofan insecticideon cottonleaves demonstrated that both specieswere susceptible to feedingon indolracarb-treated plants(Tillman et al. 2001).Few studieshave investigated the effect ofinsecticides on feeding by G.punctipes.Elzen (2001) reported that consumption ofrl. zea eggsby this predatorwas lower in malathion, profenofos, endosulfan,frpronil, azinphos-methyl,and imidacloprid treatmentscomparedwiththe untreatedconftol. Tillman etal. (2001)reportedthat indoxacarb- treatedeggs were highly toxic to femalesfeeding on theseeggs. Few studieshave compared the effect of insecticideson both TPB andG. punctipes.ln one field study, survival of G. punctipes was high for carbaryland spinosad,moderately high for indoxacarb,and low for methylparathion, lambda-cytalothrin, and imidacloprid+cyfluthrin (Leverage)(Mueggeand Payne 2001). Our research,conducted to comp.resusceptibility of;. lineolaris and G. punctipes to selected insecticideswith current or potential usagein TPB control,involved bioassaying both insects via topicaland tarsal contact routes ofadministration and evaluatingthe effect oflambda-cyhalothrinon field populationsofthese insect species.

MATERIALS AND METHODS

G.punctipes werecollected from untreatedcotton in the hill sectionof Mississippi and kept in plastic food containerswith I/. virescenseggs for food. Z. lineolaris were collected from wild hostplants in theMississippi Delta (Washington Co.), kept in cardboardice-cream cartons,and fed freshgreen beans. Immature stages were monitored biweekly in the field sothat adultscould be collectedwhen they were no olderthan l-1.5 wk old. Topical ToxicityBioassay. This testincluded the following six treatmentsand rates: (l) acephate(Orthene 75 wettablesoluable powder [0.56 kg (AI)/ha], ValentUSA Corporation, Walnut Creek, CA), (2) lambda-cyhalothrin(Karate I ernulsifiable congenhate[0.028 kg (A[)/hal, Zenoca,Wilmington,DE), (3) dicrotophos(Bidrin 8 emulsifiableconcentrate [0.56 kg (Al/hal, Novartis Crop Protection,Greensboro, NC), (4) fipronil @egent2.5 emulsifiable concentrate[0.056 kg (AI)/ha], Rhone-PoulencAgric. Co., ResearchTriangle Park, NC), (5) imidacloprid(Provado 1.6 flowable [0.053 kg (AI)/ha],Bayer, Inc., Kansas City, MO), and(6) oxamyl(Vydate2.76 concentratedlowvolume [0.28 kg(AD/ha], DupontAgricultural Products, Wilmington,DE). The recommendedrate for Z. lineolariscontol in Mississippiwas used for

４ all insecticides.A laboratoryspray chamber was used to treatadult insects topically. The spray chamberused to apply the treatmentswas equippedwith a conventionalspraying system that wascalibrated to deliver93.5 liters/ha, using a singleTX-8 nozzle(Spraying Systems, Wheaton, IL), while maintaining138 kPa pressure. The heightand speed of the nozzleabove the spray surfacewere 35.6 cm and6.4krnn1, respectively. A watercontrol was included in the test. Insectswere aspirated into a newplastic petri dish(100 x l5 mm), anesthetizedlightly (until slight knockdownor approximately3-5 sec)with CO, and then placeduncovered in the spraychamber for teatment. Control insectswere treated in the samemanner to eliminateCO, knockdownas a sourceof mortality.Before the test, a hole(55 mm in diameter)was cut in the top of thepetri dishand covered with organdymesh to increasemovement of the CO, into the dish from a CO, cylinder. A featment replicate consistedof ten insectsper species.Each treatrnentwas replicatedsix times for a total of 60 insectsper treatmentfor eachspecies, Only adult femaleswere sprayed. After spraying,the insectswere transferred to a cleanpeni dish. Sprayedinsects were provided food (greenbeans for I. lineolaris andH. virescenseggs for G. punctipes)and placed in an environmentalchamber maintained ar1)J + 2' C, 50 + 5% RH, and a photoperiodof l4:10 (L:D) h. All insectswere checked for mortality48 h aftertreatment. ResidualToxicity Bioassay. Bollgard cotton (Monsanto, St. Louis, MO) wasplanted in plots 4 (1.02 m/row) rows x 6l m andwere replicated four times.All insecticideswere applied with a spraysystem pressurized by compressedair mountedon a JohnDeere 600 high clearance sprayer.The applicationparameters were: speed - a.8 kph; pressure- 358kPa; volume - 93.4 Llha: andnozzles- TX-12 (Sprayingsystems, Wheaton, tr-). In 1996,atestwas repeated on 24, 25,and28June and included the following six treatmentswiththe samerates as intheprevious test:(l) acephate(2) dicrotophos,(3) fipronil, (4) imidacloprid,(5) oxamyl,and (6) untreated control.A randomizedcomplete block design with fourreplications was used. On 27 June1997 a secondtest was donein conjunctionwith the field studybelow andincluded the insecticide lambda-cyhalothrinat the rate in the previoustest and an untreatedcontrol. For both tests,ten cotton leavesfrom the fourth nodedown from the terminalwere collected from eachFeatment replicateforbioassayimmediatelyafterthe insecticide dried (approximately I haftertreatment). Leaveswere placed in plasticbags, transported to thelaboratory on ice,and placed in I 5 x 100 mm petri dishes.One G. punctipes and one Z. lineolariswere placed in eachof theten plastic petri dishescontaining a treatedcotton leaf. All insectswere checkedfor mortality after 48 h. Field Study.Bollgard cotton (Monsanto Company, St. Louis, MO) wasplanted in large plots,40 rows (1.02m/row) wide by 39.6m long (0.162ha), to minimizeinsect migration. A John Deere600 high-clearancesprayer equipped with a conventionalspraying system was calibratedtodeliver46.8liters/hausingTX-8 nozzles (Spraying Slntems, Wheaton,IL) and 275 kPapressure. The test began 26 June1997 and included two treatments:l) lambdacyhalothrin at 0.0128kg (AI)/ha,and 2) untreatedcontrol. A randomizedcomplete block desigrwith four replicationswas used. Sampling was done immediatelybefore each application. Post application sampleswere made at l, 3, and5 daysafter application. Samples (four rows)were taken using a KISSsampler (Beerwinkle et al. 1997).This samplingmethod was used in preferenceto other samplingtechniques to obtainsufficient insects to makecomparisons between treatments. Percentagemortality for topical and 1996residual data were convertedby arcsine transformationandthen anallzedusingPROC MD(ED followedbyaleastsignificantdifference test(LSD) (SASInstitute 2001) where appropriate. In the 1997residual and field experiments, /-testswere used for comparisonsof meansbetween treatments and species.

RESULTSAND DISCUSSION

Topical ToxicityStudy. A statisticallysignificant difference was detectedbetween species(F:4.32; df= 1,65;P=0.}aDandinsecticidetreahnents (F:117.7;df:6,65;p:

４９ 0 000 1) w hen insects w ere exposed to topical app lications of selected insecticides ぼab le l).

T A B L E l. T opical T Oxiciけ of SelCCted lnsecticidcs to a P ttα″奮 and 二.″″∽わ/is A dul低 h a Sp ray C h am ber B b ttsり , % M ortaliげ Treatment K g ぃIj/ha G. punctipes L. lineolaris

Lambda-cyhalothrin 0 .028 100 .O Jヒ0.O a, 1 100.O EL O.O a, 1

Dicrotophos 0.56 100.O JL O,O a, 1 100.O Jヒ0.O a, 1

Acephate 0.56 96 .5 ±0 ,O a, 1 94 .7 JL O .O a, 1

Fipronil 0 .056 933 と2.5a,1 93.3圭4.1ら1

Oxamyl 0 .28 83.3 ±7.l b ,2 96.7 ±4 .O a, 1 Imidacloprid 0 .053 53.3 =と15,8 c,2 66.7 =L 10 ,6 b , 1 Water O d , 1 10 ,O c, 1 a M cans w ere assessed 48 h atter treattnent,M eans w ithin a colullln and row follow ed by the sam e letter and num ber,respectivelL are not signincantly different? > 0.05;LSD ).

Both o xam yl and im idaclopridw ere less ttxic to C .P″″すっ容thtt tO二.′れ夕θわァなeven though oxam yl w as m Ore toxic than im idaclopHd tt the fom er spectes.LnidaclopHd w as the least ff e ectiv e in secticide again st ttt ri4∽わ″:s in th is top ical contact test. L am b da‐cyh alo輸直n , dicrotoph os, aceph ate, an d i pronll w ere v ery tox ic to b oth h sect sp ecies. O th er researchers h av e also rep orted th at top ical app lication s o f ottanop ho sph ates, synth tt c pyrethroid, an d i pron ll are tox ic to C ,P ″″cr″容 (L ingren an d R idBw ay 1967,T lllm m an d M ulroon ey 200 1), R 容財″α′r髄 ,ゥ s 勉ぅぇA s協宜sd cally sign in canl d ffercnce w as detccted b etw een = = = = sp ecies (F 19 .2 9 ;df 1, 12 ,P 0 .0009) an d in secticide treatm ents (F 20 .11;df ‐5, 11.9 ; = P 0 .000 1) w h en in sects w ere exp o sed to residues o f th e selected in secttcid es (T ab le 2 ). 二.

T A B L E 2 . R esidual T Ox icity of Selected lnsccticides to C .ク″″c′″容 and 五・ri4θ。ゎrなin a F ielていtreated C ot on L eaf B loassay . % M otaliげ T reatm cn t Year Kg (AI/ha) G. punctipes と.ri4∽ ゎホ Dicrotophos 1996 0.56 89.2 JL 2.9 a, 1 98,3 ±1.3 a, 1 Fipronil 0 .056 88.3 ±4 .2 a, 1 93.3 ±3.8 a,b , 1 Acephate 0.56 85.8 ±5,6 a, 1 83.2 ±5.O b ,1 Oxamyl 0.28 70.8 ±7.O b 2 9 1.7 主3.5 a,b , 1 Imidacloprid 0 .053 45,0±8.8c,2 65,0と'.3c,1

Control O d,1 4.2 d,1 ９９７ Lambda-cytalothrin 0.028 87.5±6.3ら1 87.0 と5,4 a,1 Control 0b,1 0 b,1 n Meanswere assessed 48 h afterexposure to residueson leavescollected I h afterinsecticide application.For 1996,means within a column and a row followed by the sameletter and number,respectively, are not significantlydifferent (P > 0.05;LSD). For 1997,means within row andcolumn followed by the samenumber are not sigrificantlydifferent (P > 0.05;l-test).

０ lineolaris was more susceptibleto imidacloprid and oxamyl residues than G. punctipes. However,oxamyl was more toxic to the pestthan imidacloprid.Both insectspecies responded verysimilarlytofipronil,acephate, dicrotophos, andlambda-cyhalotbrin, all ofwhichw€revery toxic to theseinsects. For G. punctipes, residual toxicity was highest for dicrotophos, fipronil, lambda- cyhalothrin,and acephate.Toxicity of oxamyl to G.punctipes was lower than that of the four former insecticides,but oxamyl was still intermediatein toxicity to this predator. For Z. lineolaris,residual toxicity washigh for all insecticidesexcept imidacloprid. Thepattem ofthe responsebetween insecticides and species to prolongedtarsal contact with dried insecticideresidues and topical contactwith sprayedinsecticides was similar for both insectspecies. Even though toxicity ofinsecticidesadministered topically appears to be higher than that oftoxicity from prolongedcontact to dried insecticideresidues, the generalpattern of toxicity betweenthe insecticidesremains similar for both methodsof application. Since the entry of these insecticideslikely would be through the exoskeletonfor both types of administration,the pattem of toxicity between theseinsecticides should be similar for both applicationmethods. Topical applications of insecticidesmay haveresulted in highertoxicity thanexposure to insecticideresidues because more ofthe activeingredient could enterthe insect with wet versusdry insecticides.Unforhrnately, quanti$ing amountsof insecticidesin treated insectswas beyond the scopeofthis padicularstudy. Residualtoxicity of imidaclopridto G. punctipesin this studywas the sameas that reportedby Boyd andBoethel (1998), a little higherthan that reported by Elzenet al. (1998), anda little lowerthan that reported by Mizell andSconyers (1992). These variances in toxicity between tests are probably due to differencesin amount of exposureof the insects to the insecticide(entire surfacetreated versus top ofleaftreated) andlevel ofcoverageon subsfates (sprayedversus dipped insecticide). Each study, nevertheless, has shown that imidaclopridis lesstoxic thanthe organophosphatesusedforTPB ccinnol. Toxicityofoxamyl residues reported by Elzen et al. (1998) was much lower than the toxicity we obtained for G. punctipes. The reasonfor our differencesis not clear,but maybedue to differencesin plantfeeding during the test. Nevertheless, we both found that oxamyl was less toxic to G. punctipes than organophosphates.Elzen et al. (1998)not only determined,as we did, thatresidues offipronil werevery toxic to G.punctipes, but alsothat they actedquickly (30 min.). McCutcheonand DuRant (1999) also showedthat residuesof acephateon cotton plants in field cages,in comparisonto residueson leavesin a peti dish,were very toxic to G.punctipes. Except for Boyd and Boethel(1998), all the residualtoxicity tests,including olus, were conductedby exposingthe insectsto residuesofthe insecticideon leavesofa plant. A problemwith this protocolis thatthe residual test may be compoundedby feedingon treatedleaves, and feeding on insecticidetreated leaves can have a detrimentaleffect on plant-feedinginsects. Geocoris punctipesfemales were susceptibleto feedingthrough dried residuesof Stewardafter sprayed onyoungcottonplants (Tillman etal.2001), To eliminateinsect feeding, insects canbe exposed to residueson the inside surfaceof a holding container.Walking on treatedleaves, though, more closelyimitates field conditions.A possiblesolution to obtainthe best of both experimental protocolswould beto preventinsects from feedingon treatedleaves. Thoughtful consideration of feedingbehavior of pestsand predators should be takeninto accountwhen conductingtests to determinethe effectoftarsal contactto residuesofinsecticides. Field Study.The daybeforeinsecticide application, the number of insectswas the same for treatedplots and unheatedcontols for both species(Fig. l). However,numbers ofinsects weremuch lower in the treatedplots comparedto the untreatedcontrols l, 3, and5 daysafter insecticideapplication for both species.In comparisonto the untreatedcontrol, a 100% reduction in TPB numbersoccurred I day after lambda-cyhalothrinapplication while a 84% reductionoccurred for thesame treatment and time after application for G.p unctipes.However, 諄 ｏ ミ リ ” ０ ０ の α 一 Ｏ． Ｅ Ｅ 口 ０ こ

13 Days after treatment

FIG' I . Meannumb ers of G.punctipes 6ndL. lineolaris per row afterapplication of cyhalothrin (0'028 kglha) on day 0. Meansbetweentreatments (untreated control, andKa.ate t eatedplots) within the same treatmentday and insect speciesfollowed by the same letter are not significantlydifferent (P > 0.05;l-test). percantage of reductionin numberfor the treatedplo'ts in comparisonto untreatedcontrols was equal(approximately 80%) for eachspecies day 3 and5 afterapplication. Thus, we concluded thatlambda-cyhalothrin hadan equallynegative impactonpopulations ofTpB and.G. punctipes in the field' Other studiesalso have shown that althoughlambda-cyhalothrin was very effeciive againstTPB in field plots,it reducedpopulations of cottonpredators, including G. punctipes (Grahamand Gaylor 1988,Studebaker 1997, Muegge and payne 2001). Acephateand dicrotophos, too, havebeen reported to be equallydetrimental to TpB and cottonpredators, including G. punctipes, in field plots l, 3, and7 daysafter treatment (DAT) @urriset al. 1986,Graham and Gaylor 1988, Studebaker 1997).1-ogiially lambda-cyhalothrin, lcephate,and dicrotophosshould be very toxic to TPB and cotton predatorsin the field since theseinsecticides are highly toxic to bothinsect species through both topical and residual routes of administration.However, fipronil whichwas equally toxic to both insectspecies in the lab, wasshown to be effectiveagainst TPB, but not detrimentalto coftonpredators, in field plotsZ DAT (Studebaker1997). Fipronil hasfeeding activity, sincemo rtality of G.punclipes females wasvery high when they fed on fipronil-treatedprey eggs(Elzen 200 I ). Thus,the reasonfor the differencesin susceptibilityof TPB andcotton predators to fipronil in thesefield testsis not "cotton apparent'It is possiblethat other speciesgrouped into predators" are less susceptible to fipronil than G. punctipes, or a reductionin tbe predatorpopulation occurredbefore the 7 DAT samplingtime. Muegge and Payne (2001) reported that imidacloprid was effective against TPB 3 and 7 DAT, but uneffective at 14 DAT at which time the population of G. punitipes droppedbelow that of the control. SinceG. punctipes wasless susceptlble to imidaclopridtlan TPB in the lab,the differencesin susceptibilitybetween the two insectspecies in the field was not surprising. G. punctipes consumesfewer imidacloprid-treatedprey than untreatedprey (Elzen2001). So the drop in populationof G.punctipes at 14 DAT may be due to loweiegg production,thus fewernymphs would developin the field or moresimply adultsdied from lack of suitablenutrition. Our topical and residualstudies indicate that imidaclopridwould be ineffectiveagainst TPB in thefield, andyet it wasreported to bevery effective in field plot tests (Studebaker1997, Muegge andPayne 2001). The insecticide mayhave feedingactivityigainst TPB increasingeffectiveness in the field. Oxamyl was very effectiveagainst TPB without reducingcottonpredators in fieldplots (Studebaker1997). Oxamyl has shown verylittle to no feedingactivity againstbrown and greenstinkbugs (Tillman, unpublisheddata), and thus, may haveno feedingactivity againstcotton predators. A possibleexplanation for the observed differencein susceptibilityoffield populationsofcotton predatorsand TPB could be that other predatorsare lesssusceptible than TPB to tarsalcontact with residuesofoxamyl as shownfor G.punctipes, and they areunaffected when feedingon cottonwith residuesof oxamyl. Further studieson feeding activity ofthese two insecticidesalong with fipronil and the effect ofthese insecticideson G.punctipes in field plotsclearly need to be doneto fully understandthe impact theseinsecticides would haveon G.punctipes populations in the field. Nevertheless,results from our laboratoryand Studebaker's(1997) and Mueggeand Payne's(2001) field studies indicatethat oxamyl and imidaclopridwould be effectiveagainst TPB while conserving populationsof G.punctipes for biologicalconfrol oflepidopteran larvae in coftonfields.

LITERATURE CITED

Bannister,J. M., G. L.Lentz, andN. B. Austin. 1995. Effrcacyof insecticideson tamished plantbugs in cotton.ESA ArthropodManagement Test. pp. 193-194. Beerwinkle,K. R., J. R. Coppedge,and T. M. O'Neil. 1997. 'KISS" - a new portable "Keep-It-Simple pneumatic Sampler"for row cropinsects, pp. 1330- 7333. InProc., BeltwideCofton Res. Conf., National Cotton Council, Memphis, TN. Boyd, M. L., and D. J. Boethel. 1998. Susceptibilityof pre.daceoushemipteran species to selectedinsecticides on soybeanin Louisian'a.J. Econ.Entomol. gl:401409. Burris, G., J. B. Graves,and A. M. Pavloff. 1986. Effectsof selectearly treatmentson flowering Junebollworms, and the early pesUpredatorcomplex. ESA Insecticideand AcaricideTest. pp. 260-261. Burris,G., B. R. Leonard,S. H. Martin, C. A. White, J. B. Graves,R. Shaw,and W. p. Scott. 1994. Fipronil: Evaluationofsoil andfoliar treatmentsfor controlofthrips, aphids, plant bugs, and boll weevils,pp. 838-844. In Proc.,Beltwide Cotton Res. Conf., National Cotton Council, Memphis,TN. Elzen,G. W. 2001. Lethaland sublethaleffects of insecticideresidues on Orius insidiosus (Hemiptera:Anthocoridae) and Geocorispunctipes (Hemiptera:Lygaeidae). J. Econ. Entomol.94: 55-59. Elzen,G. w., P. J. Elzen,andE.G. King. 1998. Laboratorytoxicityofinsecticideresidues to Orius insidiosus, Geocortspunctipes, Hippodamia convergens,and, Chrysoperla carnea. Southwest.Entomol. 23: 335-342. Elzen,G. W., andP. J. Elzen. 1999. Lethaland sublethal effects of selectedinsecticides on Geocorispunctrpes. Southwest. Entomol. 24: 199-205. Eubanks,M. D., andR. F. Denno. 1999. The ecologicalconsequences of variationin plants andprey for an omnivorousinsect. Ecol. 80: 1253-1266. Graham,L. C., andM. J. Gaylor. 1988. Effectof selectedinsecticides on tamishedplant bug andpredaceous artkopods. ESA krsecticideand Acaricide Test. pp.242. Hanny, W., C. 9. T. Cleveland,and W. R. Meredith,Jr. 1977.Effects of tamishedplant bugs (Lygus lineolarls) infestationson presquaringcotton (Gossypiumhirsutum). Environ. Entomol.6:460-462. Langston,w. c., and M. F. schuster. 1989. cotton pest control. ESA Insecticideand AcaricideT est.pp.l2l -242. Leonard, pauloff, B' R., J. B. Graves,A. M. and G. Bunis. 19g7. Evaluationof selected insecticides on tamishedplant bugs,cotton fleahopper and beneficialinsects. ESA Insecticideand Acaricide Test. p. 233. Lingren, P. D', and R. L. Ridgway. 1967. Toxicity of five insecticidesto several.insect predators.J. Econ.Entomol.60: 1639_1641. Lingren, P. D., R. L. Ridgway,and S. L. Jones. 196g. consumptionby severalcommon artbropod predatorsof eggsand larvae of tw o Heliothis speciesthat attackcotton. Ann. Entomol.Soc. Am. 6l : 613-61g. McCutcheon, G. S.,andJ. A. DuRant.1999. Survival ofselectedgeneralistpredaceous insects exposedto insecticideresidues on cotton.J. cotton. Sci. i: 102-l0ti. Micinski, ' s. 1983. Insecticidetreatrnents for tamishedplant bug and cotton fleahopper control.ESA Insecticideand Acaricide Test. p. 195. Mizell, R. F', andM. C. Sconyers. lgg2. Toxicity of imidaclopridto selectedpredators in the laboratory.Fla. Entomol.7 5: 277-2g0. Muegge, M. A., and C. Payne. 2001. Effect of selectedinsecticides on Lygus spp. And beneficialarthropods in cotton,pp. 9ll-913. Iz proc., Beltwide cotton Res.-conf., NationalCotton Council, Memphis, TN. Pankey,J. H., B. R. Leonard,J. B. Graves,and E. Burris. 1996. Toxicity of acephate, cypermethrinand oxamyl to tamishedplant bugsin vial bioassaysand caged studies on cotton,pp' 882-887.In Proc,Belrwide Cotton Res. Conf., National Cotton Council. Memphis,TN. Reed,J. T., M. s. Howell, andc. s. Jackson. 1997. Effrcaeyof variousinsecticides against tqirhed _ - - . plantbugs in Mississippi.ESA Insecticideand Acaricide Test. p. 262. Robbins,J. T., F. A. Harris, and R. E. Harris. 199g. Tarnishedplant bug control in the MississippiDelta, pp. I 197-l198.In proc.,Beltwidecotton Res. conf., Nationalcotton Council,Memphis, TN. Russell,J. s., J. Goer,K. D. Torrey,and B. R. Leonard. 199g. Evaluationof insecticide treatmentsagainst tamished plant bugs in cotton.ESA Insecticideand Acaricide Test. p.244. SASInstitute. 2001. sAS/c onlinedocumentfr, Release g.2,cary,Nc, sAS Institute,Inc. scott, w. P., andG. L. snodgrass.1996. Efficacyof Fipronil andprovadoagainst tamished plantbugs in cotton.ESA ArthropodManagement Test. p. 26g. Shaw,R., andH' S.Yang. 1996.Performance summary of Fipronilinsecticide on cotton1996, pp. 862-865.In Proc.,Beltwide cotton Res.conf., Nationalcotton council, Memphis, TN. Snodgrass,G. L., and G. W. Elzen. 1995. Insecticideresistance in a tarnishedplant bug populationin cottonin the MississippiDelta. southwest. Entomol. 20:317-323. Studebaker,G. 1997. Efficacy ofselectedinsecticides on plant bugsand predatory arthropods on cotton.ESA ArthropodManagement Test. pp. 272-273. Teague,T. G., andN. P. Tugwell. 1996. Efficacyof provadowhen appried with feedgrade molassesto conEoltamished plant bugs.ESA ArthropodManagement rest. pp.169- 270. Tillman, P' G., and J. E. Mulrooney. 2001. Effect of malathionon beneficialinsects. Southwest.Entomol. Supplement 24: 13-21. Tillman, P. G., G. G. Hammes,M.. Sacher,M. Connair,E. A. Brady,and K. Wing. 2001. Toxicityofa formulationofthe insecticideindoxacarb to thetarnished plant big, Lygus lineolaris Qremiptera:Miridae), andthe big-eyedbug, Geocorispunctip"-s luemiptera: Lygaeidae).Pest Management Science 58: 92-100.

４ V O L .2 8 N O .1 S O U T H W E ST E R N E N T O M O L O G IS T M A R .2003

IN S E C T S A S S O C IA t t W I「H B L A C K S T A IN R 0 0 T D IS E A S E C t t S IN P IN Y O N P IN E S T A N D S

R. JasonBishop and William R. Jacobi

Departmentof BioagriculturalSciences and Pest Management ColoradoState University, Fort Collins,Colo. 80523

ABSTRACT Black stainroot diseasecenters in pinyon-juniperwoodlands near Cortez, Colorado were monitored in 1998 and 1999 for potential insect vectors of the fungus that induces black stainroot disease(Leptographiurn wageneri var. wageneri). Baited Lindgren funnel and pitfall traps were placed in root diseasecenters from mid May to Decemberin 1998 andmid May until early Augustin 1999.Insects representing four ordersand 28 families wer€ captured includingl3 bmk beetle species(Scolytidae) and 13 weevil species (Curculionidae).More than 50Voof the scolytid and curculionidspecies were root and stump breeding insects. These insects could be vectors of the pathogen becausethe fungusresides in roots andlower stemof pinyonpine trees.Most scolytidswere trapped in May and June; whereas,the majority of curculionids were trappedin July and August. T\ree Hylastesspecies were the most cornmonlyand consistentlyrecovered beetles or weevils. Furtherstudies are neededto determineif any of the insectsare vectorsof the black stainpathogen.

INTRODUCTION Pinyon pine (Pinusedulis Engelm.), occupying more than 22.5 million hectaresin the westemUnited States,plays an important role in southwestemecosystems (Mitchell and Roberts 1999). Pinyon pine occurs with Junipentsspp. in mixed woodlandson foothills, low mountains,mesas, and plateausfrom 1,200 to 2,400 m (Bucknranand Wolters 1987). Pinyon pines have died in unprecedentednumbers in southwestern Coloradoover the last decade@ager 1999). Sustainedmortality of pinyon pine has negatively affected recreational sites such as campgroundsand picnic grounds and the value of residentialhousing. The largenumbers of deadand dying treeshas increased tuels for wild fires @ager1999). The two primary mortality agentsin the region are black stainroot diseasecaused by the fungal pathogen,Leptographiwn wagenei (Kendr.) Wingf. var. wageneri and /ps bark beetle,Ips confusus(I€Conte). Black stainroot diseaseis a vascularwilt disease (Landis and Helburg 1976) and was first recognizedand describedfrom observationsat Mesa verde National Park in 1942. when a pinyon pine is killed by black stain root disease,the fungus is found in the outer annual rings of roots and lower stem. t-ocal spread of the pathogen occurs through root contacts and grafts between infected and healthytrees (wagener and Mielke 1961). Ips bark beetlesoften attackpinyon pines affectedby black stain root disease./ps bark beetlescan also kill pinyon pine without the involvement of black stain root disease(Keams 2001). In ponderosapine (pinus ponderosaLaws.) and Douglas fir (Pseudotsugammziezii (Mft.) Franco.) ecosysterns 的謂:需品盤t鑑ず1樹itts塩崩瑞 麟喚程襴概譜寸品s齢 さ描,噛絆 滞鮒:;颯私総&洲品黙。紺i辞魅艦品渋解艦 fr°m the suspected hsects to徒助航ne whch hsect sedes tttuallve :鮒8慶骨. p y ctor the l M □T H O D S ° htt N撤艦 、艦 淵 浮盤雰と艦ポ桃盟 協齢群転登農i鮭 襴 路欄見盟群鮮獣艦軽端号権轍景鑑 ぐ 脚 ヘ

traps w ere not used in 1999 because the pitfali traps collected sind lar num bers alld species as the ttght traps ぼlg. 1). Further, it could not be det銅 ned w hether the i l d ch orvos insectcidc in the L indgren funnei traps w as effectve in retttning allinsects.

総 襴 び 縄 強 襴 穏 烏 1 ■sects captured w ere separated and identifled to the farよly level at C olorado State University.Only curcu1lonids ttd scolガdS Were further idendied since insects in these BrOupS Vector the p athogen in other tree spectes, Identin cations of thc curculionids w erc

臨 襴 欄 艦 盟 麒 辮 羅 ゾ 鵠:肥鉛賢培謂歯F船 提ど&盟盟控∬と i絆撒 艦 All specimensare on deposit in the C.P. Gillette Museum of Arthropod Biodiversity at colorado StateUniversity' REsuLTs AND DlscussloN We identified insectsrepresenting four orders and 28 families (Iable 1) including 13 bark beetle species(Scolytidae) and 13 weevil species(Curculionidae) (Table 2). More than 50 % of the trappedbark beetleand weevil speciesare known to feed or breed in roots or stumps(Amett. 1971,Fumiss and Carolin 1977,O'Brien and Wibmer 1982, StephenWood, Brigham Young University, Provo Utah, personal communication). Thus, 12 or moreinsects we foundcould be potentialvectors.

TABLE 1. InsectFamilies Trapped within Black StainRoot DiseaseCenters in Pinyon PineStands near McPhee Reservoir in the SanJuan National Forest in 1998and 1999.

Flemiotera Coleootera Hvmenootera

Reduviidae Carabidae Bostrichidae Meloidae Asilidae Ichneumonidae Lygaeidae Staphylinidae Anobiidae Cerambycidae Calliphoridae Pompilidae Coreidae Histeridae Trogossitidae Chrysomelidae SarcophagidaeVespidae Pentatomidae ScarabaeidaeCleridae Curculionidae Tachinidae Buprestidae Ciidae Scolytidae Elateridae Tenebrionidae

Potential Scolytidaevectom include; Dendroctonusvalens (LeConte),Hylastes fulgidus (Blackman), Hylastes gracilis (Irconte), Hylastes macer (kConte), Gnathotrichus denticulatus (Blackman), Hylurgops porosus (Leconte), Hylurgops reticulatus (Wood), Otthotomicuscaelatus (Eichhoff), Orthotomicuslatidens (kConte) and Xyleborus intrusus (Blandford). Only two Curculionidae, Ortorhynchusovatus (L.) and Otiorhynchusrugosostriatus (Goeze) representpotential vectors. The most commonly trapped beetle, /ps confusus,was not assumedto be a vector becauseit is rarely seen breeding within a meter of the ground or in roots of pinyon trees. Ips confususwere active in the region during the collection period, which explains the high trap catches. Hylastes spp. and Hylurgops porosus were the most abundant of the potentialvectors, More potentialvectors may occuramong the specieswe trappedbut the life histories of many of theseinsects are not known. In fact, the life histories of most of the specieslisted as potential vectors are not well known. Potential vectors were recoveredfrom both pitfall and flight traps. Both scolytids andcurculionids were trapped from May to Novemberin 1998and from May to August in 1999. With the exceptionof I. confusus,the majorityof scolytidswere trapped in May while the most curculionids were trappedin July and August (Frg. 1). The Hylastes spp. werecollected in both traptypes but predominantlyin the pitfall traps. Despitethe fact that an insect has yet to be implicated as a vector of the pathogenof black stainroot diseaseof pinyonpine, someevidence suggests a role for insectvectors. Landis and Helburg (1976)reported insect galleries, presumably made by bark beetles, adjacentto stain margins of I'eptographium wageneri var. wageneri in diseasedpinyon pine roots. Researchand observations on otherhosts of theblack stainpathogen suggests a vectorrelationship in pinyon pine. Goheenand cobb (1978)reported Ieptographium wagenei var.ponderosum (Haringlon and cobb) conidiophoresand peritheciaonly in insectgalleries of diseasedponderosa pine roots. As insectsare well-known vectorsof va4ouslzptographium utd Ophiostomapathogens, Goheen and Cobb (1978) suggested that insects might introduce lzptographium wageneri to ponderosa pine roots. Harrington et d. (1985) and witcosky et al. (1986) demonstratedthe ability of the

７ ‐ 境 爵豊代S ヽ吟= ヽョ ^R--ogSgnoa;

ヨ ' ヽ 1 , 1

ａ 費 ０ 9 , S , 1 , 目 ∞ 験 せ “ ０ 目 パ 「 ∞ 〓 【 “ 堂 〓 劇 “ あ 縁 ５ 目 ｂ 旬 資 畳 宮 Ｌ 目 吾 召 ０ ｂ ‐ い ， Ｓ じ ０ 脅 将 Ｓ ｚ 旨 目 掌 卓 , ‐ 畳 ‐ 1、 1 , 1 , ぃ 畳 罵 ｂ ∽ 協 営 「 芭 担 Ｓ 填 ｙ ８ ８ 宮 ８ . 望 1 ｇ Ｘ 】 Ｏ Σ ８ ０ 車 十 ８ 芝 ， 卜 ， ヨ 紀

８ 彗 ギ

＆ 写 悩 至 】 ミ 中 崎 ' 1 1 l ， 豊 い い Ⅲ 露 中 ‐ ミ 増 置 頂 ● ぁ ロ 営 宮 ョ ゴ 呂 旨 呂 tt Nl_ ,1 ， ‐ ‐ く ０ se ｔ 招 ■ ｎ 芭 雨 ま セ Ｂ ミ し ■ 遵 き Ｒ 遭 ３ 査 的 も 食 （ 習 以 重 ご ミ Ｓ ミ ミ Ｓ ミ ０ 電 ヽ ミ お 営 ミ ヽ ミ や 富 ｇ ョ ミ ゆ 増 ミ ミ Ｓ 送 ミ Ｓ Ｅ ミ ヨ ゆ せ Ｓ ヽ ０ 漫 く ミ も 卜 崎 習 や へ ミ ミ ｓ 高 ヽ ゆ ヽ ヽ ミ Ｓ せ ミ ヽ ヽ ち ヽ ミ 豊 Ｌ 爵 ヽ 一 営 】 惣 ミ 望 終 ミ る ヽ セ ミ 口 宙 や ヽ 【 ０ Ｓ Ｓ ミ 的 革 ヽ 起 思 控 鷺 中 、 馬 ミ ミ さ ｓ も 】 臼 ミ ミ 望 豊 ふ 曽 豊 ミ Ｓ ヽ ヽ Ｓ 車 塩 豊 噂 買 車 寒 ● 買 吟 母 玉 車 き Ｓ 津 Ｏ ヽ 定 お 母 Ｓ 書 ミ ミ ミ 、 や ョ ョ「 も 畳 ヽ ミ ミ 空 蜜 ヽ 母 ミ 、 的 も ミ 出 旨 超 超 Ｓ ミ 、 Ｓ 革 蛸 ｚ ミ ミ 車 ミ 営 ミ 世 Ｓ Ｓ 串 資 ミ ミ ゆ 登 ミ や ＆ 資 も 章 望 ｏ ヽ Ｓ ち ミ 喧 串 望 車 袋 』 国 ミ ヨ ｔ 曽 Ｓ 連 ヽ 望 ミ ミ ヨ 、 ヽ Ｓ 桜 ０ Ｓ や ヽ 単 ヽ Ｓ 一 日 辞 的 Ｓ ミ ｇ ミ ミ】 ミ Ｓ ｆ ふ や 蛍 竜 ミ 電 さ 登 貸 沢 尽 Ｓ 増 富 ミ ぺ Ｓ ぷ ミ ヽ Ｓ ま Ｓ 民 ュ ミ 暮 ミ ち ｔ ミ や や ミ 的 貿 Ｓ せ 、 Ｓ ■ る ミ や ミ 出 ８ じ、 Ｆ ヽ ｐ ０ ミ や Ｓ Ｓ や ヽ」 Ｓ ｐヽ だ い ３ ｓ 崎 ゆ ０ ｏ “ 草 臣 Ｎ 逗 ， ｏ ｏ ・ ｏ ，

５ S colytidS C aptured 1998 ｍ ｍ の ０ ｍ 一ま 伸 田B pitfa li trap る ０ 卸 囲Lindgren ttnnel の キ 獅 ｍ 口total 。 ″ 3/t2 8/29 1177

ScolytidsCaptured 1 999 ∞ の □ ptfa ‖tra p ０ ５０ 一ま 石 ∞ Lind re n ttn ne l 秘 □ g ５。 (nOt uSe d) ＋一 。 口tota l

CurculionidsCaptured 1 998 の ０ 一 20 Ｃ 回 p 【fa H tra p Ｏ 15 〓 ョ せ 10 □ Lind g re n ttnne l ョ ０ 5 E]tOta l 荘 0 5/20 6/19 11″ 12115

C urc ulionids C aptured 1999 ~ ~ ~ 弔 国pttitrap

圏日ndgren ttnnel (nd uSed) 口 to ta l

７ ２ 確 Ｄ 範

H G . 1. N um bers of adult Scolytids and C urculionids by date that w ere trapped near M cP hec R cservoir in the San Juan N ational F orestin 1998 and 1999.

９ scolytidHylastes nigrinus (Mannerheim), pissodes and curculionid, fasciatus(l*conte) and steremnius cainatus (Mannerheim) to vector L. wageneri var. pseudotsuage (Hglq- and cobb) in Douglas-fir. Adult beetles mide feeding'wounds and artificially contaminatedadults successfully transferred the pathogento rJ"ang wounds. Additionally, at leasttwo insectsof thesegtnera consistently appeir in diseasedDouglas- ponderosa !l -"nd pine stands. Either recovereddirectly i.otn uff""teO trees or Eaps; Hylastes and,Pissodes are reportedto constituteupwards of gOVoof insectsrecovered in theseareas (Witcosky and Hansen 19g5, Jacobi l9O2). We haveidentified a groupof bark beetlesand weevils that might be vectorsof the black pathogen. stain The next step will need to focus on deterriining if any of the potential vectorsidentified in this study are vectoringthe black stain"pathogen.The biology of theseinsects in regardsto pinyon pine and black stainroot diseasealso will needfurther elucidation. A careful collectionand rearing of insectsfound in roots of decliningand recentlydead pinyon pine affectedby black stainroot diseasewould also help explainthe long distancetransmission of the blackstain pathogen.

ACKNOWIEDGMENT

we jhank Holly Keams and sam Harrison,and Boris Kondratieff,Department of BioagriculturalSciences and Pest Management, Colorado State University, for assistance with field work and insect identificationrespectively, and rom Eager,USDA, Forest service, Forest Health Management,Gunnison service center and San Juan National Forest,for suppliesand information during this study. Scolyids were identified by StephenWood, Brigham Young Universityand curculionidswere identifiedbv Charles W. O'Brien,Florida A&M Univenitv.

LITERATURECITED Amett, R. H. Jr. 1971.The beetlesof the united states.American Entomol.Inst. Ann Arbor,Michigan. 1112 pp. Buckman,R. E., and wolters, G. L. 1997. Multi-resourcemanagement of pinyon- juniper woodlands. In Everctt,R, L. Ed, proceedings-pinyon-Juniperconierence. USDA ForestService, Gen. Tech. Rept. INT-215 Eager,T. J. 1999. Factorsaffecting the healthof pinyonpine trees(pinus edulis)in the Pinyon-Juniperwoodlands of westerncolorado. .ln proceedings: Ecology and nranagementof Pinyon-Junipercommunities within the interior west. compiledby s. B. Monsenan R. s. Stevens.RMRS-P-9. ogden,ur. usDA, ForestService, Rotty Mountain ResearchStation. pp. 397-399. Fumiss,R. L., and carolin, v. u. tgtz. westem forest insects.USDA ForestService, Misc.Pub No. 1339.654 pp. Goheen,D. J., and cobb, F.w., Jr. 1978. occurrenceof verticicladiella wagenerii (black strain root diseasein Pinw ponderosa) and its perfect state, Ceratocystis wagenerisp. nov., in insect(Dendroctonus valens). phytopathology 6g: I 192-1195. Harrington,T. C., Cobb, F. W., Jr., and lownsbery, J. W. 1935. Attiulty of Hylastes nigrinus,a vectorof venicicladiellawageneri, in thinnedstands of Douglas-fir. can. J. For. Res. 15,'519-523. Jacobi,W' R. 1992. Potentialinsect vectors of the black stainroot diseasepathogen on southernVancouver Island. J Entomol.Soc. British Columbia g9:54-56. Kearns,H. s. J. 2@1. Black stain root diseasein the pinyon-juniperwoodlands of SouthwesternColorado. MS Thesis.Colorado State University. tZO pp Landis, T. D., and Helburg, L. B. 1976. Black stain root diseaseoi pinyon pine in Colorado.Plant Dis. Rep. 60:713-717.

６０ Mitchell, J. E., and Roberts,T. C., Jr. 1999. Distribution of pinyon-juniperin the westemUnited States.1z S. B. Monsenand R. Stevens.Eds., Proceedings: Ecology and Managementof pinyon-juniper communities within the interior west. USDA ForestService, Proceedings. RMRS-P-9. O'Brien, C. W., and G. J. Wibmer. 1982. Annotated checklist of the weevils(Curculionidaesensu lato) of North America,Central America, and the West Indies(Coloeptera: Curculionoidea). Mem. Amer.Entomol. Instit. No.34:l-382. Wagener,W. W., andMielke, J. L. 1961. A staining-fungusroot diseaseof ponderosa, Jeffrey,and pinyon pine. PlantDis. Rep. 45:831-835. Witcosky, J. J., and HansenE. M. 1985. Root-colonizinginsects recovered from Douglas-fir in various states of decline due to black-stain root disease. Phytopathology75 :399 402. Witcosky, J. J., Schowalter,T. D., and Hansen,E. M. 1986. Hylastes nigrunus (Coleoptera:Scolydidae)Pissodes fasciatus, urd Steremniuscarinatus (Coleoptera: Curculionidae)are vectors of black stain root diseaseof Douglas-fir. Environ. Entomol.l5: 1090-1095.

６ V O L .28 N O .1 SO U T H W E ST E R N E N T O M O L O G IST M A R .2003

ARTIFICIAL FEEDING SYSTEM FOR TI{E SQUASHBUG, ANASA rflSz.t @E GEER) (I{ETEROPTERA:COREIDAE)

Blake R. Bqrtiner , Astri Wayadandet,B.D. Bruton2,S.D. Pail, ForrestMtchell3, and JacquelineFletcherr

ABSTRACT

Squashbugs, Anasa /rislis (De Geer) (Heteroptera:Coreidae), did not feed on liquid diet-filled feedingsachets used traditionally for feedinghomopterans, or on pouch- like artificial feeding source(AFS) containinga meridic diet developedfor rearingthe western tarnished plant bug, Lygus hesperusKnight, another heteropteran. However, excisedcubes ofsquash fruit, vacuuminfiltrated with a suspensionofthe desireddiet and offeredabove a screenbarrier, were aocepted. During a 48-hrtesting period, all cube-fed adult insectssurvived, while 35% of thosefed on the meridic diet offered in parafilm pouches and 40o/oof those fed on 5% sucroseoffered in sachetsdied. After blue food coloringwas addedto the dietsas a marker,the excretoryfluids of 75% of insectsfed on infiltrated cubeswas blue in color indicatingthat feedinghad occurred,while no blue defecationoccurred when dye-amendeddiets were offered in the other two systems. The A. tistis AFS provides a convenientsystem for studying the feeding behavior ofsquash bugs and provides an alternative to whole plants for studying the etiology of cucurbit yellow vine disease.

INTRODUCTION

The squash bug, Anasa rrisfis (De Geer) (Heteroptera: Coreidae), occurs throughoutNorth Americaand is considereda majorpest of cucurbits.A. tristisfeeding involvespiercing ofthe plant's epidermisby the styletsand intracellularpenetration to 'lacerate the mesophyllor vasculartissues @eard 1940, Bonjour l99l). This and flush' method of feeding intemrpts xylem transportof water due to vasculardamage and blockagethat resultsin collapseof plantstructures distal to feedingsites (Neal 1993). Cucurbit yellow vine disease(CYVD) is characterizedby rapid and general yellowing of leavesappearing over a 3-4 dayperiod, followed by gradualor rapid decline and death of the vine in severalcucurbit crops (Bruton et al. 1998). A. tristis was reported(Pair et al. 2000)and confirmed @extine et al., unpublisheddata) to be a vector of Serratia malcescens,the causalagent of CYVD (Bruton et al., in press). This bacteriumhas been cultured from diseasedfield-grown watermelon, zucchini, pumpkin, andcantaloupe.

l D epartm ent of E ntom ology and P lant P atholo8y,O klahom a State U niversi呼, Sj llw ater, O K 74078,U SA K 7455乳U SA 粘 着 鞘 闊 盟 掛陥 鑑 拙 革締 静 潮 品 i 梢茸 Ⅲ P resent address of senior authort D epartm ent of E ntom ology, U niv of C alifornia, R iverside,C A 92507 Although artificial - -feeding systems(AFS) have been developed for many homopteraninsects including leafhoppers,aphids, and whiteflies, toiw have been developedfor heteropterans(cohen 2oo0b). AFSs,along with artificial diets,havb been developedfor the westerntarnished planf bug, Lygus iesperus Knight rconen .2000a; Debolt patana 1982; Debolt and 1985), thJ reduviid big, Triatoina in/estans Klug (schaub l99l), and the southem green stink bug, Nezara iiriauu (L.) (Ragsdale* al. 1979). In the latter case,the AFS was instrumentalin associating'insectrlahg wittr microorganismtransmission. The . relationship betweenthe vector A. tristis and the pathogenS. marcescensis an important elementof the etiology of cyvD, but investigationiof pathogen-vector interactions.were hamperedby the lack of a convenient-AFs uy wtrictr-pathogen acquisitionby the squashbug could be assuredand controlted. preliminaryanempts to feed l. tristis on known homopteranor heteropteranartificial diets were unsuccessful. The goal of this studywas to developan AFS and a suitablediet on which to maintain squashbugs during researchstudies and to providea convenientacquisition source for the microorganism.Feeding acceptance and survivorship of A. rzisrison traditionalAFSs were comparedwith thoseon a new A,FScomposed of diet-infiltrated squashcubes.

MATERIALS ANDMETHODS

colonies of A. rristis,initiated with adultscollected from a field nearLane, oK, were maintainedin screenedcages (50 x 25 x 45 cm) in a growth room at 27"c, 12L:12D, and 45-50%oRH. colony insectswere rearedon purptin plants(Cucurbin "connecticut pepo L. var. pepo Field") that were replacedat- weekly intervals and supplementedwith washedzucchini squash fnrit (C.pepoL. var.melopefo). SeveralAFSs and artificial dietsused for-sustaining other hemiptlran insects were offered to A. trislis and observationsof feeding were made. Feeding sachets,or stretched-parafilmmembrane feeding systems, commonly used for studyi;g leafhopper, aphid,and whitefly feeding(Mitsuhashi and Koyama l97l), hereafterrerened to as the lromopteranfeeding system, consistedof 29.5m1flexi-cup medicine cups (Baxter Heathcarecorp., Deerfield,IL) coveredby one layer of parafilm (AmericanNational can, Greenwich,cr) stretchedto 4X its original sizeover the openingof the medicine cup. A volume of 0.5m1of 5olosucrose (pH 7.0) was placedon the outer surfaceof the stretchedparafilm and coveredwith anotherlayer of stretchedparafilm to enclosethe solution (Mitsuhashiand Koyama l97l). Twenty insects,not separatedby sex, were placedinto the AFS, one per sachet,and the sachetswere maintained at 27oi, l2L:12D, and45'5o%o RH. Anotherfive insectswere individually offereda modificationof this membranefeeding system in which squashslurry, made by lique$ing approximatelyl0g squashfruit and 20ml steriletap water in a blender,was brushedonto the insect-facing surfaceof the membrane(lml squashslurry per l6cm2membrane surface) as a possibli gustatorycue. A non-membranesystem was also iestedby offering semi-solidpreparations of 5%oagarose or 3ologelatin. Ten ml of solutionwas pouredinto the bottomof eachol'five, 60 x l5 mm petri dishes@agesdale et al. l97g). A 70x20mm, fiberglassscreen barrier with 1-mm grid squareswas placedso that it restedon the rim of the petri dish. Five insects(sex not determined)were placedon top ofthe screenin eachdish, and the dish lids were positionedon top, their rims restingon the screen.The disheswere maintained for 48 hr at27oC,constant light, and45-50% RH. Intact squashfruit were washedwith handsoap and reverse osmosis (RO) wateq and subsequentlywas cut with a sterilerazor blade. The epidermiswas removedand 6- mm' cubeswere excisedfrom the cortex. Squashcubes were vacuuminfiltrated with blue dye so evidenceoffeeding could be determinedby evidenceofblue dye in squash

御 bug excretory fluid. For infiltration, squash cubes were submerged in RO water coitaining blue food coloring(McCormicli and Co., Inc, Hunt Valley,,MD) (O'lml per tOml HzO) at a rate of I cubeper ml in a 500-mlErlenmeyer vacuum flask.. A vacuum force was applied for 5 sec, during which negativepressure forced air from the intercellular ipaces of the tissue, and as the vacuum was quickly released the blue- coloredwater enteredthe intercellularspaces. Adult squashbugs were confined individually with a squashcube using an apparatussimilar to that?escribedby Ragsdaleet d. (1979). The insect-wasplaced in th" botto. half of a 60 x 15 mm petri dish. A 70 x 70 mm squareof sterile l-mm fiberglass screenwas placed over the dis[ the squashcube was positioned in the center ofthl screeq andthe petri dish lid was positionedon top. The squashcube was offered abovethe insect,rather than below, to avoid excretoryfluid contaminationofthe cube' The feeding apparatuseswere incubatedin a humidity chamber(placed on a stand in a coveredtransparent plastic shoeboxcontaining water at a depth ofapproximately lcm) to minimizecube desiccation and were held under constant light x27"C. Three of the AISs (the homopteranmembrane feeding systeq the tarnishedplant bug feeding system, and ihe newly developedinfiltrated squashcube 4pS described ab6ve; amendedwith identical ratios of blue food coloring were tested for A, tristis acceptanceover a 48 hr period at 27oC,constant light' and 45-50ploRH. In the homopteran membranefeeding system, 20 adult A. ttistis were placed individually into the apparatusesand given accessto 0.5m1of 5olosucrose (pH 7 0) containing'bluefood coioring(0.lml per loml sucrose)(Mitsuhashi and Koyama l97l). Tf,e tarnishedplant bug feedingsystem (Cohen 2000a) consisted ofmeridic diet, a lima bean meal and wheai germ basedformulation containing additional nutrient supplementsand containing blue food coloring(O.lml per l0ml diet) enclosedin l00mm * Zir-. unstretchedparafilm pouches as describedby Debolt and Patana(1985). This diet was offeredindividually to 20 singleinseots (not separatedby sex),with the parafilm plates 35-mmwells' -pouches supported on meshscreening in six-welltissue culture with The A. tristis feeding systemwas also offeredindividually to 20 A. '/t'rtis (not separatedby sex). For all threi feedingsystems filter paper(Whatman no. 4), cut to fit the bottom of the apparatus,allowed collection of the droplets of squashbug excretory fluid. If this fluid was blue, squashbugs were assumedto havefed on the offered diet. Insectmortality over the 48-hrperiod was recorded. for boih detecrionofblue excretoryfluid and insect mortality, chi squareanalysis wasused to assessdifferences (critical P-value=0.05) between treatments (SAS 1996).

RESI.JLTSA}ID DISCUSSION

In our preliminaryexperiments, when squashbugs were placedin severalAISs and offered artificial diets known to be acceptableto other homopteranand heteroperan insects,neither stylet insertionnor test probingwas observed.These insects did not feed throughany ofthe testedmembrane based feeding systems, nor would they feed on the semi-solidagarose or gelatinformulations. Squashbugs ied on blue, water-infiltratedsquash cubes within one hour of introduction,with 15 of 20 test insectsexcreting droplets of blue fluid onto the filter paper after 48 hr (Table l), whereas none of the insects offered the other two systemVdietsexcreted any fluid (chi-square=40,dF-'A, P<0.001). None of the twenty insects offered blue water-infiltrated cubes in our AFS died during the 48-hr period, comparedto 7l2Oof those offered the meridic diet in unstretchedparafilm bags and 8/20 of those offered 57o sucrosein feedingsachets. Thesepercentages were significantly different amongtreatments (chi-square= I 0. I 3, dF2, P<0.0 I ). t_ ６ 鮒熱撚齢 襴 襴 鰍 i路麒ぢ i鮮 T A B L E l Preference andsuitability of ThreeDifferent Artificial FeedingSystems by the Bug,Anan tristis. a B lu e E xcetory F lu id 4 8 1r Treatment" B lu c N ot B lue Su rvi D ied Syttem w嗣 15 2 0 亀総鯛静 0 卯 SyStem with M eridic 0 e 2 0 8 12 artiflcial diet F eed ing Sachets w ith 5% 0 2 0 7 13

隅 酬 鮒 漁 鮒 襴 脇 雛 枇 uare lα呼 乳 チ練麓欝罷縄麟軽罷鰹鑑品‰韓 °群齢淵:S七 b 播艦in仏胡;鵠ご チ st t r鍋ヽし 器

６６ be expectedin the plant. Becausethe CYVD pathogenis phloem-associated,testing S. ^*"it"r^ transmiisionby A. tristisin the artificially inoculatedsquash cube AFS may not exactly mimic acquisition or inoculation of the pathogenin a naturally infected plant system. The vacuuminfiltration of a squashcube saturatesall intercellular spaces,.which would not occur in plant infection, and doesnot assurebacterial entry into phloem sieve tubes. However, ttie AFS developedfor A. tristis provides a tool that will facilitate the studyof CYVD.

ACKNOWLEDGEMENTS

This researchwas fundedby the USDA-SouthernRegional IPM Programand the Oklahoma Agricultural Experiment Station. Thank you to Edmond Bonjour and KristopherGles for reviewingthis mamrscript.

LITERATURE CITED

Beard,R.L. 1940.The biology of Anasa'ristis Deceer. Bull. 440 Conn.Ag. Exp. Stat.: 595-682. Bextine, B.; A. Wayadande,S. Pair, B. Bruton, F. Mitchell, and J' Fletcher' 2001. Parametersof Serratiamarcescens transmission by the squashbug, Anasa tristis. PhytopathologyJune, 2001. 91 (6 Supplement):S8, Bonjour,E .L., W,S. Fargo,J.A. Webster,P,E. Richardson,G.H' Brusewitz. 1991. Probingbehavior comparisons ofsquash bugs (Heteroptera: Coreidae) on cucurbit hosts.Environ. Entomol. 2O:143-149. Bruton, 8.D., F. Mitchell, J. Fletcher,S.D. Pair, and A. Wayadande,U. Melcher, J. Brady, B. Bextine and T.H. Popham.2002. Senatia marcescens'a phloem- colonizing, squash bug-transmittedbacterium, is the @usal agent of cucurbit yellow vine disease.Plant Dis. (In press). Bruton, B.D., J. Fletcher,S.D. Pair, M. Shaw,and H. Sittertz-Bhatkar.1998. Association of a phloemJimitedbacterium with yellow vine diseasein cucurbits.Plant Dis. 82:512-520. Cohen, A.C. 2000a.New oligidic productiondiet for Lygus hesperusKnight and Z. lineolaris(Palisot de Beauvois). J. Entomol. Sci. 35:301-310. Cohen, A.C. 2OOOb.A review of feeding studiesof Lygus spp. with emphasison artificial diets.Southwest. Entomol. Sup. 23: I I 1-l 19. Debolt, J.W. 1982.Meridic diet for rearingsuocessive generations of Lygus hesperus, tamishedplant bug.Ann. Entomol.Soc. Am. 73:119'122. Debolt,J.W., andR. Patana.1985. Lygus hesperus.Amsterdam: Elsevier. Mitsuhashi,J., andK. Koyama.1971. Rearing of planthopperson a holidic diet. Entomol. Exp.Appl. 14:93-98. Neal, J.J. 1993. Xylem transport intemrption by Anasa frislis feeding causesCucttbita pepoto wilt. Entomol.Exper. App.69:195-200. Pair, S.D.,B.D. Bruton,F. Mitchell, andJ. Fletcher.2000. Yellow vine management,pp. 145-148. ht Proc. lfth Ann. Hort. Ind. Conf., J. Motes (ed). OklahomaState University,Stillwater. Ragsdale,D.W., A.D. Larson,and L.D. Newsom.1979. Microorganisms associated with feeding and from various organsof Nezaraviridula. J. Econ. Entomol. 72:725- 731. SAS. 1996. Version6.12. SASInstitite Inc., Cary,NC. Schaub, G.A. l99l Pathological effects of Blastocrithidia triatomae (Trypanosomatidae)on the reduviid bug Triatoma infestans after infections by membranefeeding and long-term stawation. J. Invert.Path. 58:57-66.

６ A R .2003 V O L .2 8 N O .1 SO U T H W E ST E R N E N T O M O L O G IST M

LONGEVITY OF ULTRA-LOW-VOLUME SPRAYSOF FIPRONIL AND MALATHION ON COTTON IN ME)(ICO

J. E. Mulrooney,K. A. Holmesl,R' A' Shawl'and D' Goli2

APTRU,USDA-ARS, P' O' Box 36, Stoneville'MS 38776

ABSTRACT

In 1996, fipronil and malathion residueswere evaluatedafter four ulta-low-volume applied at r*t applications in northeasternTamaulipas, Mexico' Sprays were Ct-t"r/i at 840 g A.I'lha' b.88 i/h;. fipronit ** uppn"a at 28 and 56 g A.L/ha and malathion intervals' Leaf Four applicationswere mai! bginning r,r May at.four' five and six day residuesof surfaceresidues of malathion acJumuto:teOwitheach application' Leafslrface nproiif uppfi"a at both ratesdissipated >90% after 2 to 4 d after all applications.

INTRODUCTION

(uLV) insecticides in crop Probably the most extensive use of uhaJow-volume -1978' From production-hasbeen in the Boll weevil EradicatiohProgmm which beganin great The il;-;;;ng, ULV application of insecticideshas been used to effectiveness. cost of the i"U"t i"A nion"v .u*i by using ULV application has helped to keep the with ultra- pi"gd "t an affordable level, V-ast*teug" can be most expeditiouslyfeated and less io*lnoi*"r of insecticidesbecause aircraft are able to spendmore time spraying time frlling and ferrying to and from the aintip' programs fechnicat malatlion is the insecticide of choice for area-wide eradication It is uguin.t the boll weevil, Anthonomusgrandis grandis Boheman,in the United States' eFfective--Fi;;ii;n"g*ttasainst this insect as an ulta-low voiume (ULV) spray(Jones et al. 1996,1997)' from Aventis EnvironmentalScience, Research Triangle, NC) is 1 demonstratedto be effective as a high phsoif'rri"ii"'rp*v pytu;le (boliet et al. lgg2) that has been at 0.056 kg A.I./ha uganrt the botl weevil in field plots (Bunil et al' 1994)' L/ha four appiicationsof ULI sprayso1 malathion and fipronil in cottonseedoil at 0.88 were eifective against-insecticidethe boll weevil in a field test (Reedet d' 1998)' factor in iow long an remainseffective on the plant ry*face f an_imp.ortant d*erminin! the optimum spray interval for contolling boll weevils. Costly over-treating can be avoidedif the longevigof the insecticidebeing usedto confol a pest is known. The objective of this-test was to compar€the longevity of ULV spraysof fipronil and malathion on cotton leavesafter multiple applications'

Aventis Environ. Science,2 T.W. AlexanderDr. ResearchTriangle Park,NC 27709. 2 rive W ilson N C 27 96. soudlem Testm g and R esearch Labs.,Inct,3809 A irpo■D , , ,

６９ M ex icO)w as tested at 840 g A .I./ba. A pplic筑10ns Of each tream cnt w ere m at 継盤繍鱗鯛 宙配お°配 協 盆鷲1軒紺ぷ置獄艦掛盤掛寮批就盟穐齢 繊措岳∬ 苺 議潜

L eaves w ere sam pled o, 1, and 2 d aFter tht 鞄轟麒縄軽鑑 nprottandmalathittwttwashedttmmeuαand Wers pp tt Waceofeachにポushg Drted in rettgerated cOntainers tO the U SD A 辮路部!檻罵鞘盤 盟骨相襴 療 ペ 血総習鰍 温盈!撤d靴淵 諮品雛 : 襴撒告艦絆ポ襴 i r 柵 路 糊 テ 品:檻lぐ :.材盤 品ま ofdetectiOn w as O.12 ppm . s斜 部 辞鵠S稗器品 艦 ぺ on upp er and low er sw faces Of leaves an路cr ea 鉛 w ere collected al ler the sとst,second ,and fOIHth app licationst 札群(襴 町弼 鰍 sts o f r t en d s (in tCrcep t an d slop e) and slop es 餓 灘 e.

０ RESULTSAND DISCUSSION

Analysis ofvariance showedsignificant differencesin residuesofmalathion and high and low rates of fipronil between days (F=5.0;dF3,3;P<0.02)and applications(F: 10.53;dF2,3;P<0.0001) Malathion residueson the upper surfaceof the leaf rangedfrom 398 to 4,391 nglcm' and from 61 to 1367 ng/cmzon the lower surface(Table l). After the first application, residueson upper or lower surfacesof leavesdid not significantly decreaseuntil two days after application. After applicationstwo and four, there were no significant differencesin residueson either the upper or lower leaf surfacesbetween day zeto and day two sample dates. Malathion residueson the lower swface significantly decreasedfrom day two to day fow after applicationstwo and four, A ULV spray of technical malathion at 2.8 kg A.I./ha near Brownsville, Texas (Wolfenbargerand McGan 1971)deposited 6,490 ng malathion/cm'on leaves. In our study, the firsJ and fourth applicationof 0.84 kg A.I./ha deposited2,105 and 3,381 ng malathion/cmzon leaves(Table 1). The 2.8 g A.I/ha of malathionapplied in Brownsville, Texas, resultedin 2- to 3- fold greaterdeposition than the 0.84 g A.I./ha rate used in our study.^Residuesof malathion^onupper leaf surfacessampled on day zeroafter one (1,621 n{cm'), two (2,168 n!cm') and four (2,576 ng/cm') applicationsshow increasing residuesafter multiple applications. Residuescollected immediately after applicationstwo and four increased35 to 600/ofrom applicationone. Residuesat 2 d after applicationstwo and four were about 10 fold greater than application one. There was no significant diflerence in residueson the upp€r leafsurfacebetween 0 and 4 d after applicationfour. Malathion residueson the upper and lower surfacesofthe cotton leavesincreased after the first and secondapplication (Table 2). Interceptsofregressions ofresidues after the secondand fourth applicationwere equal. The slope of regressionof malathion residue on the upper leaf surfacewas negative afterapplicationsoneand twoand positive after applicationfour(Table2). This result seemsto indicate a decreasein the rate of degradationof malathion on the leaf surfaceas the seasonprogressed. In contrastto malathion,residues of fipronil on the upper surfaceof the leaf on day zrro did not accumulateafter eachofthe three applications(Table l). Fipronil is a short-lived insecticidecompared to malathion. At both rates,residues of fipronil decreasedan average of 95%o2 and4 d after the three applications. Residuesof the low rate of fipronil on the upper leafsurface decreasedsigrrificantly on eachofthe daysafter application. Residues of fipronil on the lower leaf surface decreased98o/o 2 d after the first application at the high rate and 94Youtd93% 4 d after the secondand fourth application, respectively. Residuesof fipronil were not sigrificantly different on days0 and I after the first application. The residues found on the lower leaf surface at the high rate after applications two and four were significantly different from each other on each of the sampledays. Fipronil residuedeclined on both surfacesof the leaf after all applications(Table 3). The loss of fipronil was always greateron the upper surfaceof the leaf than on the lower surface. The intercept, which is an estimateof the initial deposit on the leaf, of the high rate of fipronil was alwaysgreater than that for the low rate. The interceptdecreased from applicationone to applicationfour for the low rate of fipronil, especiallyon the upper surfaceofthe leaf. At the high rate, the interceptwas variable on both leafsurfaces. The intercept of fipronil at the high rate far exceededthe low rate after applicationstwo and four.

７ 田 ６ ｏ ０ つ や ｏ 寸 ｍ 吟 ∞ い 寸 卜 ． 崎 「 ⇔ 〇． ． ベ． い ヘ 【 督

あ 田 「 つ つ ０ い 苺 ０ ０ と へ い 寸 崎 い ． ． 一 ∞ 〇 ば い ． ． じ や い 崎 ∞ 口

宮 ｅ

車 車 、 ち ゆ Ю 「 Ｇ 『 ミ 〇 〇 卜 ， ． 崎 卜 ゆ 崎 ゆ い 〇 ， ｏ 崎 へ 【 寺 い． い 鷺 ０ ∞ い り 来 ｏ 営 ｏ 母 Ｏ ｏ ｏ つ つ ｏ 旦 】 寺 い い 寺 い い い い 。 ． ∞ 。 ｏ あ 卜 ０ 【 ｍ 【． ぶ 日 】 督 易 ｏ ． ＞ ｏ 舎 Ｓ Ｓ つ Ｓ Ｑ ｍ ０ つ 卜 つ ０ い Ｈ い 崎 い ∞ 一 い ふ い 〇 ． ． ｍ ． 崎． 【 寝 寺 い 崎 い 【 ０ 樽 【 ■ せｓ

む 割 ｏ 房 Ｅ ｏ 】 Ｏ 田 ｏ 、 母 車 、 田 ｏ あ ａ 翻 ∞ ゆ い 寺 へ ヨ Ｅ Ｄ Ｏ ゆ ョ ゆ や 〇． ∞． 卜 寺． ぶ 留 一 い 卜 ∽ ∞ い ∽ 増 い 【 女 く 副 ｂ 営 ま の Ｅ Ｅ ） 封 ｏ ， ｏ つ ロ 」 崎 ｏ ｏ ０ ｏ つ 電 」 Ｎ ∞ ゆ Ｏ 一 ∞ い 申 一 い ． ． Ю 〇． 。 ｏ ∩ 留 い 【 【 ｏ 中 ■ 亀 Φ 宙 ａ 洋 ＞ 壇 ｏ 、 苺 車 一 忠 卜 ０ つ へ つ 卜 罵 環 一 へ ∞ 〇 章 ． 首 定 ∞ ， ． 崎 ． 寸 日 卜 ∞ へ い 【 Ｉ で 一 ・Ｓ 日 Ｒ Ｅ 単 史 あ ｏ 聾 Ｇ Ｓ 、 雨 、 Ｓ 浄 Ｏ こ ” 崎 ∞ Ю ヽ 鷺 〇 【 寺 総 へ い， 卜， ∞ 崎． Ｎ． 一 Ю Ю 崎 寸 【 べ ｂ 分 声 【 ０ ＞ ｏ あ ｏ ９．

９ 〓 営 路 】 ） ｏ 一 ｇ 】属 ● 屈 汽 『 苺 苺 ｄ 連 ”も ミ ミ ミ 導 輯 ｏ 的 的 的 的 ヨ の 樹 も Ｆ 〇 〇 〇 ｏ 】 ． ． ． ， Ｐ ぁ 電 ｇ ∞ つ 。 ∞ ｃ 電 ． ｏ へ い 。 Ｎ 崎 β 【 Ｐ 【 一 「 【 】 、 口 「 召 召 壇 旧 ▼ ｏ Ｓ お お Ｓ Ｈ 雨 、 ｅ ｅ 、 ｅ ｅ 琶 臼 営 ， Ｂ ３ 電 ３ Ｐ Ｓ Ｒ く 母 】 停 ＞ Ｌ Ｌ 吉 Ｌ Ｌ Σ 材〕 ■ 【 ｏ S I ' 撒是F養緋 櫨tt 鮒燃曲 :粋盤l側『智ゴ縄陥

UDDer Srace

1 2080 ‐1.12

2 2893 -0.45

4 2864 0 ,12 Lo w er Surface

1 567 ‐1.77

2 1075 ‐0 .36

4 1022 ‐0 .6 1

= ate T A B L E 3 . E quatio ns は eSidue Intercept X D ザ ) of F iprod l R esiducs from U pper and L ow er S r aces of c ot on L eaves as a F un cd on of D ay (A pp ttcaよon on D ay O). Rate (g A.L/ha) Application Intercept Slope Upper Surface 28.0 1 67 -3 .5 1 56と1 1 68 ‐3 .79 28,0 2 30 ‐2 .13 56.1 2 80 -2 .55 28.0 4 l3 ‐2 .24 56,1 4 4t ‐2 .56 Lower Surface 28.0 1 l6 ‐2 .85 56.1 1 36 ‐3 .2 3 28.0 2 l5 -1.60 56.1 2 49 ‐1.71 28.0 4 9 -1.12 56.1 4 4l -■71 TABLE 4' contrasts of Trends and Slopesfor Low and High Ratesof Fipronil on upper andLower SurfacesofCotton Leaves. S10De r8 1) C onm st

A pp licatiOn F v alue F value

U pp er s urface

1 0.65 O.5225 0.55 0。46 14 2 4 .82 0 ,0333 1.86 0.169 4 7.56 0.0073 1.22 0 .2722 L ow er S u acc 1 3.83 O .02 56 0.68 0.4 106 2 15.57 0.000 1 0。13 0.72 18 4 3.32 0.0574 2.69 0,1049 p6: intercept,p1 : slope.

M alath iOn h as been efFectively used in the eradication of the boll w eevil. Its efFectiv cn ess along w ith its low co st w ill llkely en sure its cOntinued u se in erad ication .

繍 盟 含辞 鑑 襴 鞠 盟 鞘 麒 艦 撤 縦 accom plished w ith flprOnil providcd that it w as iabelcd for use on cOtton in the U .S .

職 S堪 ;靭 齢 緊 盟 紺 1艦 縄 機 器 鑑 騨 坪

A C K N O W L E D G E M E N T

４ LITERATURE CITED

Burris, 8., B. R. Leonard,S. H. Martin, C. A. White and J. B. Graves.1994. Fipronil: evaluationof soil and foiliar treatrnentsfor control of thrips, aphids,plant bugs and boll weevils, p. 838-8,M. 1z D. Herber and D Richter [ed.] proc. ol ttre geittride cotton Insect Researchand contol conference, san Diego, cA, 1994. National Cotton Council, Memphis,TN. carlton, J. B. 1992. Simple techniquesfor measuringspray deposits in the field--u: dual sideleaf washer.ASAE PaperNo.921618. Am. SocAgric. Eng.,St. Josep[ MO. colliet, F-, K. A. Kukorowski,D. w. Hawkins,and D. A. Roberts.tilz. ripronil: a new soil and foliar broad spectruminsecticide. Brighton Crop Protection Conf"."rrce- Pestsand diseases, p.29-34. BCPC, Croydon, UK. Harmon, N., R. Shaw, and H. Yang. 1996. worldwide development of fipronil insecticide,p. 759-765.12P. Duggerand D. Richter [eds] proc. Beltwide cotton Researchand contol conference, New orleans, LA, 1996. National cotton Council,Memphis, TN. Jones,R. G., D. A. wolfenbarger,and osama El Lissy. 1996. MalathionuLV ratestudies under boll weevil eradicationprogftrm field conditions,p. 717-719. 1n p. Dugger and D. Richter [eds] Proc. Beltwide Cotton Researchand Control Conference,Niw Orleans,LA, 1996.National Cotton Council, Memphis, TN. Jones,R. G., and D. A. wolfenbarger.1997. MalathionULV residualcontrol efficacy underboll weevil eradicationprogram conditions, p. 1208-1209..lz p. Duggerani D' Richter [eds] Proc. Beltwide Cotton Insect Researchand Control Conference, New Orleans,LA,1997. NationalCotton Council, Memphis, TN. Mulrooney, J. E., and D. Goli. 1999. Efficacy and degradationof fipronil applied to cotton for control of Anthonomusgrandis grandis (coleoptera: curculionidae). J. Econ. Entomol.92:13641368. Mulrooney, J. E., K. D. Howard, J. E. Hanks,and R. G. Jones.1997. Applicationsof ulta- low volume malathion by air-assistedground sprayerfor boll weevil (coleoptera: Curculionidae)contol. J. Econ. Entomol. 90:639-645. Mulrooney, J. E., D. A. wolfenbarger, K. D. Howard, and DeepaGoli. 199g. Effrcacy of ultralow volume and high volume applicationsof fipronil againstthe boll weevil. J. CottonSci. 3:104-1 10. Reed,D. F', M. christian, w. G. Jones,and D, w. Kiser. 199g. Aerial applicationof fipronil (Regent) vs. malathion in a replicated field test for boll *eevil, p. 1262- 1264. In P. D,,gger and D. Richter [eds] proc, cotton Insect Researchand control Conference.San Diego, CA, 1998. NationalCotton Council, Memphis, TN. sAS Institute. 1997. sAS/srAT software: changesand enhancementjthrough release 6.12. User'sGuide. SAS Institute, Cary, NC. wolfenbarger,D. A., and R. L. McGan. 1971. Low-volume and ultra-low-volumesprays of malathion and melhyl parathion for control of three lepidopterousconon pests. USDA-ARSProduction Research Report No. 126,14 p. V O L .28 N O .1 S O U T H W E ST E R N E N T O M O L O C IST M A R .2003 SCIEW IFIC N O電

INJURY AND DISTRIBUTION OF ONION THRIPS (THYSANOPTERA: THRIPIDAE) IN RED CABBAGE HEADS

Tong-Xian Liu and Alton N. Sparks, Jr.l

Texas Agricultural Experiment Station, Texas A&M university, 2415 E. Highway g3, Weslaco, TX 78596-8399 rPresent Address: Department of Entomology, coastal plain Experiment Station, p. o. Box 1209, University of Georgia, Tifton, GA 31793

onion thrips, Thrips tabaci Lindeman, has been a serious pest of onions in south Texas for decades(Royer et al. 1986, sparks et al. 199g, Liu and chu, unpublished data). z tabaci can causesignificant damageto crops adjacentto onion fields shortly before, during or shortly after onion harvest between March and May becauseof the dispersal and migration ofhigh thrips populations from the onion fields. Direct feeding by larvae and adults of z. tabaci not only causesdirect cosmetic damage of the cabbagehead with rough brown blisters on green or white cabbages, but also significantly reduces storage duration and market value (Fox and Delbridge 1977, Stoner and Shelton 1988, Shelton et al. 199g). In the spring 2000, we observed that a sevdrely infested red cabbage field adjacent to an onion freld near Mission, Texas. The cabbageswere planted in mid-December 1999. on l0 March 2000, we collected 50 red cabbageplants from this field. we carefully cut the plants with all leaves at the ground level and placed them in plastic bag for hansport to the laboratory for examination. Before assessinginjury and thrips on the cabbage heads, we rernoved all old wrapper leaves. We then peeled offindividual leaf layers from each cabbage head, numbering respectively from the outmost toward the center of the head. We visually evaluatedthe injury level ofeach leafbased on the following four categories:0: no injury; = 1 1-5 small injured spots or blisters; 2 = >5 small injured spotsbut still somewhatisolated; : and 3 general injury across much of the area covered by the leaf. At the same time, we counted all thrips (larvae and adults) on both surfaces of each peeled leaf from each cabbage head. we stopped peeling off leaves when no injury or tlrips were found from five consecutive leaves of the cabbagehead. We collected the adult thrips and identified them under a microscope. We found that all thrips collected were T, tabaci. We did not identify larvae, but assumed that all of them belonged to the same species. voucher specimens were deposited in the Insect Collection of Texas A&M University Agricultural Research and Extension Center at Weslaco. The correlation between leaf iqiury and number of thrips found on that leaf was atalyzed using pROC CORR (SAS Institute 2001). on green or white cabbage, the feeding damage symptom caused by T. tabaci adults and larvae appeared as bronze discolored blisters or spots on the leaf(Fox and Delbridge 1977, Stoner and Shelton 1988). However, the blister-like spots were even more distinct on the red leaves than on green or white cabbageleaves. Therefore, the cosmetic damage and market value were even more significant for red cabbagethan for green or white cabbage. The whole field sampled was destroyed without harvesting.

７ 回 の 「 の ≡ 」 〓 」 I Injury ｏ Nl Thrios ５ ｃ ｏ」 石 ＞ 聖 と コ 官 一

.10 z 3 4 5 6 7 I 9 11 12 13 14 Leaves(from the outmostto the centerof the head)

FIG. 1. Injury and number of r. tabaci (lawae and adults) on leavesof red cabbagehead (n = 50).

The irgury levels from the outmost older leaf of the head, with wrapper leaves previously removed, to the younger leaves toward the center of the head are shown in Fig. l. Most damageoccurred on the outer older leaves.The injury level in each of the fust three layers of leaves was >2, considered as severe damage. The fourth and fifth leaves had moderate damage with an injury level of l.l and 1.6, respectively, and the next six inner leaves had little damage.No injury was found from the twelfth and younger leaves of the head.Most heads examined would have required removal of the wrapper leavesand at least five outer leaves from the head in order to reduce damage to an acceptablelevel, which would have reduced head size and marketability to an unacceptablelevel. Consequently,the level of damage in this field and similarly infested fields generally results in complete loss of the crop. Similar to the injury level of the leaves,more thrips were found on the outer leaves ofthe heads than on the inner or younger leaves (Fig. I ). The outer four leaves had g I .3% ofthe total thrips, and the outer eight leaves 98.0% ofthe total thrips. An average of

LITERATURE CITED

Fox, C. J. S., and R. W. Delbridge. 1977. Onion thrips injuring stored cabbagein Nova Scotia and Prince Edward Island. Phytoprotection 58:57-58. North, R. C., and A. M. Shelton. 1986. Colonization and intraplant distribution of Thrips tabaci (Thysanoptera: Thripidae) on cabbage.J. Econ. Entomol. 79:219-223. Royer, T. A., J. V. Edelson, and B. Cartwright. 1986. Damage and conhol of Thrips tabaci Lindeman on spring onions. J. Rio Grande Valley Hort. Soc. 39:69-74. SAS Institute.2001. SASiSTAT users' guide,Version 8.01,Cary, NC, USA, p 1028. Shelton, A. M., W. T. Wilsey, and M. A. Schmaedick. 1998. Management of onion thrips (Thysanoptera: Thripidae) on cabbageby using plant resistanceand insecticides.J. Econ. Entomol. 9 | :329 -333. Sparks, A. N., Jr., Anciso, J., Riley, D. J. and C. Chambers. 1998. Insecticidal control of thrips on onions in south Texas: Insecticide selection and application methodology. Subtrop.Plant Sci. 50:58-62. stoner, K. A., and A. M. Shelton. 1988. Influence of variety on abundanceand within-plant distribution ofonion thrips (Thysanoptera:Thripidae) on cabbage.J. Econ. Entomol. 81:1 1 90-1 195. Ｆ

VO L .2 8 N O tl SOUTHW E ST E R N ENTO MOLOG IST MAR.2003 SC IEN T IFIC N O TE

RANGE EXTENSION, NEW HOSTS, NEW MORPHS, AND A SYNONYM OF THE SEDGE APHIDI RHOPALOSIPHUM MUSAE (SCHOUTEDEN)

StephenW. Taber

Campus Mail 1004, St. Edward's University 3001 South CongressAve., Austin, Texas

During a study of the aphid genus Rhopalosiphum Koch which culminated in a phylogeny for this economically important group of insects (Taber 1994), a population of one species was discovered where it might never have been suspected. The species was Rhopalosiphum musae (Schouteden), and the population was found in western Maryland at orchard Pond, Allegany county, Greenridge state Forest on 23 November 1991. This aphid previously was known in the United States only from the western part of the country and from no closer to the Mid-Atlantic Statesthan the Rocky Mountains, a distance of roughly 1,500 miles. The species has also been reported in the far west of Canada,in British Columbia (Forbes and Chan 1983). The Maryland host plant was a sedge or bulrush of the genus scirpus (cyperaceae). This genus is the only known summer host of a seasonally migrating aphid that uses cherry, almond, and other Prunus speciesas winrer hosts (Gillette and Palmer l932a,b; Patch 1938; Palmer 1952; Richards 1960; Smith and Parron 1978). Winged adults, wingless adults (Fig. 1), nymphs' and parasitized mummies were clustered in large numbers on bulrushes near the waterline' With dropping autumn temperatures the plants were dying back rapidly and the insects were destined to follow suit. Microscope slides borrowed from museum collections yielded three previously unreported life cycle stages and two previously unreported host plants. The stages or morphs are the fundatrix and spring migrant (from unidentified host, Ash Creek, Utah, l- "BM V-1934, coll. G. F. Knowlton, British Museum slide 1984-340 Rhopalosiphurn scirpifuIii = rnuse€", det. D. H. Ris Lambers) and the fundatrigenia (ftom Prunus fasciculata, San Bernardino Co., Cal., 10 May 1968, coll. R. C. Dickson). Photographs of all three morphs are available but not published here due to spacelimitations. One previously unknown host plant is the native desert almond Prunus fasciculata upon which the fundatrigenia morph was found, and the other is the exotic bird-of-paradise or crane flower Strelitzia reginae ("Hawaii via California via Florida: West Palm Beach"; Florida State Collection of Anhropods). The ornamental bird-of- paradise is a close relative of banana for which the aphid was named when it was first described from Belgian greenhouse specimens on leaves of Musa ezsere (Schouteden 1906). Occurrence on banana was reported again decades later (Hughes and Eastop l99l). A study of this material corroborated the opinion of David Hille Ris Lambers that Rhopalosiphum scirpifulii Gillette and Palmer is a junior synonym of R. musae' Voucher specimenswill be sent to the United StatesNational Museum in Washington, D.C.

l H ett ptcra A phididae

Palmer, M. A. 1952. Aphids of the Rocky Mountain Region. Thomas Say Foundation 5: r-452. Patch, E. M. 1938. Food-plant catalogue of the Aphidae of the world, Part V. Bull. Maine Agric. Exp. Sta. 270:47-100. Richards, W. R. 1960. A synopsis of the genus Rhopalosiphum in Canada (Homoptera: Aphididae). Can.Entomol.92 (Suppl.13): l-51. Schouteden,H. 1906. Catalogue des Aphides de Belgique. Mem. Soc' Entomol. Belgique 12:189-246. Smith, C. F., and C. S. Parron. 1978. An annotated list of Aphididae (Homoptera) of North America. Tech. Bull. No. 255, North CarolinaAgric. Exp. Sta.: l-428. Taber, S. W. 1994. Labile behavioral evolution in a genus of agricultural pests: the Rhopalosiphurn plant lice (Hemiptera: Aphididae). Ann. Entomol' Soc. Amer. 87:311-320. |

V O L .29 N O .1 SOU THWESt t R N EN T O MO L O G IST MAR .2003 SCIEN T IFIC N O TE

RANGE EXTENSION, HABITAT, AND REVIEW OF THE RARE ROBBER FLY' )RTH OGONI S STYGIA (BROMLEY)

Stephen w. Taber2and Scott B. Fl"eno.t

While conducting a biodiversity survey in the Ottine Swamps of south central Texas (Gonzales County), we captured a remarkable inchJong robber fly that was unfamiliar to us despite years of field work in the wetlands and forests of this area. It was identified as Orthogonis srygia (Bromley), the only Nearctic speciesof a mainly Old World genus (Hull t962, Ir4artinand wilcox 1965, Josephand Parui 1981, Poole 1996), with the aid of the standard key to Nearctic Asilidae (Wood 1981). This is the first published report from the southwestern United States. Our single specimen is a female; the male sex remains unknown to science. Furthermore, the biology of the species, even ils regards such basic information as habitat, is unknown. Our purpose here is to provide basic biology, to alert others to search for the missing male, and to provide the first illustration of O. stygia to ud those wishing to continue our work (Fig. 1). An intriguing line drawing of the wing may be seen on page 558 (Fig. 14) in wood (1981). It bears the male symbol and if that accurately described the specimen at hand, then paradoxically the known female has never been illustrated in any way though the unknown male has. O. stygia has been reported previously only from the southeasternUnited States. The roughly half dozen reported specimens are from Florida (Gainesville), North Carolina (Stovall), and Mississippi (ovett) (Bromley 1931, 1950). Thus, more than half a century has elapsed since the last published encounter with this striking predator that mimics a spider wasp with its black and metallic blue coloration. The Texas A&M University Insect Collection in College Station has two additional females. Both were collected by Bromley himself in 1934 in Liberty, Texas, and both are shown here beneath our own specimen. Bromley's site is 160 miles east and slightly north of our collection locality. We found the asilid on a trail in an ash swamp characterized by dwarf palmetto, Sabal minor (Jacquin) Persoon, and green ash, Fraxinus pennsylvanica Marshall. These plants are more typical of the southeastemU.S., where the fly has been previously reported on two occasions, than of central Texas or the southwest, and the palmetto does reach its western limit nearby as does the loblolly pine and presumably O. stygia (Maxwell 1970, Schultz 199?). Our specimen and all those reported in the literature except one were captured in June. The single exception is one of Bromley's Texas flies captured on 2 July 1934. Prey, oviposition site, and larval habit remain to be discovered. We confirm the "undoubtedly statement made in the original description of the species,that the species is very rare" (p. 434, Bromley 1931). Yet we modify the statement that O. stygia is confined to the southeastern United States. Its absence from a report on the Asilidae of Texas (Bromley 1934) is explained by the discovery of Texas specimens in the same year 'Diptera:Asilidae 'Campur Mail 1004,St. Edward'sUniversity, 3001 South Congress Avenue, Austin, Texas 78704-6489 '305 w 35th Street#204. Austin, Texas78705-1443

Schultz,R. P. 1997. loblolly pine: The ecologyand culture of loblolly pine (Pinustaeda L.). AgriculturalHandbook 713, USDA Forest Service, Wash., D. C. Wood,G.C. 1981.Asilidae,pp.549-573.lnManualof NearcticDiptera,Vol. 1-.J.F. McAlpine et al. [coordinators.]Research Branch, Agric. Canada,Monograph No. 27, CanadianGov. Pub.Centre, Hull, Quebec,Canada. vol".28 NO.2 SOUTHWESTERNENTOMOLOGIST JUN.2003

Brol,ocy oF coLEoTHORpA DOMINICANA FMNCISCANA (LBCONTE)'

J. E. Slosser

TexasAgricultural ExperimentStation P. O. Box 1658,Vernon, Texas 76385-1658

ABSTRACT

The life stagesof Coleothorpadominicanafranciscana (LeConte), a case-bearingleaf beetle, are discussed,and photographsof each stageare presented. Eggs are oviposited in clustersnear the tip of spineson the prickllpear cactus,opuntia engelmanniiSalm-Dyck, and the femalebeetle constructs a unique,previously undocumented, funnel structureon the spine to protect the eggs. Larvaeare active for 5 - 8 months,and the overwinteringstage lasts about 6 months. Developmenttime from eggto adult was oneyear in the laboratory. This clytrine is associatedwith the antFormica neoclara Emery, but therewas no evidencethat the ant was necessaryfor lawal development.

INTRODUCTION

The biology of Coleothorpadominicanafranciscana (Leconte) is poorlyunderstood, eventhough information on the generalbiology ofthis case-bearingleafbeetle was reported as early as 1874(Riley 1874). Eggs are coverpdwith plateletsof fecal material that are molded aroundthe egg,as described by Riley (1874),Erber (1988), and Stiefel and Margolies (1998). Rather than being dropped to the ground as occurs with many clykines, each egg of C. dominicanais attachedto one endofa shortstalk that is fixed at the otherend to the oviposition substrate,an arrangementsimilar to lacewingeggs (Neuroptera: Chrysopidae). In his review of the biology of case-bearingleaf beetles(Camptosomata), Erber (1988)reported that many speciesin the subfamilyClytrinae were totally or partly myrmecophilous,and Riley (1891) indicated tlwt C. dominicana S.) was associatedwith the arftsFormica obscuripesForel, Formica sP., and Camponotussp, Eggsof severalclyfines are carriedby antsinto their nests ( Erber 1988,Stiefel and Margolies 1998)where the eggshatch. The larvaeremain in the ant nest and develop into adults. After hatching, the larvae live within the egg case for the remainderofthe larval andpupal stages.As the larvagrows, it usesfecal material, fine sandand small piecesof vegetativematerial to enlargethe case. Clytrinae adults are pollphagousand feed on leavesfrom treesand bushes, frequently without a definite preference(Erber 1988). Thereared beetles in this studywereidentified using the current taxonomy of Moldenke (1970), a work that beatsthe Norttr-AmericanChrysomelidae of the subfamily Clyhinae. The presentgeneric placement of this beetle(in Coleothorpavs. Coscinoprera)follows Moldenke (1981). Most specimensof Coleothorpadominicana franciscana (kConte) from the southwestemUnited Statesdiffer in their outward appearancefrom thoseoriginating from the t Coleoptera:Chrysomelidae: Clytrinae

9l moreeastern portion ofthe subspecies'range. The southwesternspecimens, including theadults rearedin this study,closely match the type specimenof Coscinoptera dorsalrs LeConte (images of LeConte's tlpe specimencan be viewed in the Museum of ComparativeZoology type specimansdatabase, at http:llmcz-28l68.oeb.harvard.edu/mcztypedb.htn).Coscinoptera dorsalisis presentlyconsidered ajunior synonymof Coleothorpadominicana franciscana. Coleothorpadominicanafranoscana occurs in TexasRolling Plainsrangelands, but its biology from this regionhas not beendocumented. This articlereports general life historyof this speciesand provides photographs ofeach life stage.

MATERIALS AND METHODS

The study areawas located4.0 krn north of Knox City, Knox Co., Tex., on the eastside ofHighway6atthejunctionwithapavedcountyroad.Duringtheyears 1990to 1995,thesite was consistentlysearched for eggsand adultsweekly from early April to mid-June. Sporadic observationswere made in 1989and in 1996-1998. When eggswere found, they were takento the laboratoryat Vernon, Texas. Attempts were made to rear the larvae from the eggs every year. However, only two were reared successfullyto adult in 1995-96. Lawae were maintainedin a 85-mm diameterpetri dish; the lid had a 15-mm diameterhole in the middle that was coved with fine-meshscreen to allow ventilation. The filter paperin the dish was moisteneddaily with a drop of water. The rearing arenawas cleanedevery-other-day, and small piecesof freshly diced cactuspads, small pieces of moldy bread,several smearsof pinto beandiet (Shoreyand Hale 1965),and two, l-cm lengths of dental wick soakedr^ l0% honey water were placed on the filter paper as food sources.Also, small amountsof fine sandand driedplant materialwere addedto aid the larvae inenlargingtheircases.Rearingwasconductedinthelaboratoryunderroomconditions[4:10 (L:D) photoperiod,temperature 24-27' Cl. The two larval caseswith the last instar larvaewere transferredfrom laboratoryroom conditionsinto an environmentalcontrol chamberon 29 November 1995. Photoperiodin the charnberwas11:13(L:D),andtemperahrewas18.3o:15.6"C(L:D).On4April 1996,the larvaVpupalcases were retumedto the laboratoryroom, and on 22 Apilthey were transfened to an environmentalcontrol chamber maintained at 26.7"c with a 14:10(L:D) photoperiod.

DEVELOPMENTAL STAGESAND BIOLOGY

Eggs. Eggsare ovipositedin clustersnear the tip of a spineon pricklypearcactus, opuntia engelmannii Salm-Dyck. Each egg is attachedto the cactusspine by a short, thin filament. The lengthof the eggstalk was not measured,but Gilbert(1981) reported the length of the stalk was about5 mm in Coscinopterapanochensis Gilbert. The ridgesand depressions of the molded fecal material covering the egg give the impressionof a small pine cone. Egg clustersare preceded on the spineby a previouslyundocumented, funnel-shaped structure which completelyencircles the spine (Figs. la lb). The funnel appearsto be composedof the same fecal platelet material that coversthe individual eggs. The pointed end ofthe funnel facesthe egg cluster while the open end of the funnel facesthe cactuscladophyll (pad). The funnel is reminiscentof the rat b.oi"r rhi"ldr utilized on a ship's mooringlines *d -uy function to protectthe eggsfrom predators.The funnelswere present with six of eight eggclusters; two of the six funnelswere collapsed around the spine,perhaps due to faulty constructionor to damage by beatingrains. The earliestoviposition was recordedon 8 May andthe lateston 2 June. The datesthat eggclusters were found andnumbers of eggsper cluster(inparenthesis) were: 8 May 1989(17), 20May l99l (21),l2May 1993(12),24May 1995(19), 2 June1995 (la), andthree clusters

92 FIG' L Life stagesof Coleothorpadominicanafranciscana (LeConte):1a. egg clusterwith funnel; lb. eggswith filamentattachment to spine;lc. younglarva feedingon moldy bread; ld. maturelarva (lcm long); le. overwinteringcase with sealedend; 1f. adult.

% on8Mayl997(18,ll,andl8). Gilbertreporteduptol3eggsforC.pcnochensis.Eggclusters held in the laboratoryat room temperaturehatched in 5-9 da1o. Ten ofthe 33 eggs(9 ofthe 19 egg-group,and I ofthe 14egg-group) did not hatchin 1995;otherwise, all eggshatched'in the other years. Larvae. If the arenabecame too wet, a larva would exit its case,or it might chew and deshoythe case,and eventuallydie. ln someyears, smashed ants or other insectswere added to the petri dish. However,the only time that larvaofed readily on insectremains was in 1993 when ground-uppieces of Formica neolcaraEmery were provided,but this behaviorwas not consiste,nt.Mesquite leaves cut into small pieceswere freque,ntlyprovided, but the larvaedid not appearto feed on this fresh material. Lanraewere observedto feed consistentlyon mold growing on both the bread(Fig. lc) and cactuspieces, and larvaewere frequentlyfound under or on top ofthe honey-soakedcotton wicks. The larvae did not feedon pintobean diet in 1997. Molds appearedto be a nutritional requirement,but the larvaefrequently became enhapped in the mold anddied. Riley (1874, I 882)successfully reared C. dominicanalarvae on deadand decayingleaves. In 1993one larva remainedactive throughlate January1994, and two larvaeremained active until mid-Decernberin 1997 before dying. In 1995, the two larvae that survived to becomeadults sealed their larval casessweral times during the surnmerand fall, apparentlyto molt. The caseswere sealedfor the final time duringOctober. The activelarval stagelasted 5-8 months. A casecontaining a maturelarva was ca. I cm long (Fig. 1d). OverwinteringS/age. The overwinteringstage (note the sealedflat endofthe casein Fig. le) lastedabout six months. The timing of occurrenceof the pupal stageis not known because thelarvaVpupalcaseswerenotdisturbed.Erber(1988)indicatedthepupalperiodlasted12-28 days. Adults. Oneadult @ig. lf) emergedon 3 May andthe otheron l0 May 1996. Adult longevity was approximately56 days. Adult anergencein the laboratorycoincided with the timing of oviposition in the field, but adultswere neverobserved at the collecting site, Under laboratoryconditions, the completelife cycle from eggto adult was one year. However,and it may havebeen coincidence, eggs were only found every-other-yearfrom 1989to 1997,which implies a two-year life cycle. Pinnedspecimens identified as Coscinopteradominicanafranciscana LeConte, in the collection at Texas Tech University, Lubbock, TX, were collected 4 June 1970 @ig Bend NationalPark, Brewster Co., Texas, Chisos Basin) and 3 April 1969(San Angelo, Texas, kion Co.). Smith andUeckert (1974) collectedCoscinoptera sp. from mesquiteflowers andpods in Dickens,Lubbock,andLynnCountiesinwestTexas.SpecimensofCoscinopt*aaeneipennis LeConte andC. axillaris LeConte,in the TexasTech University museum,were also collected from mesquitein west Texas. RangeSite. Availableevidence indicates C. dominicanafranciscanais associatedwith rangelandhabitats dominated by honeymesquite,Prosopis gland.ulosaTorr.,innorth andwest Texas. The oviposition site on prickllryear spinesand constructionof the protective funnel suggestsa preferencefor slenderobjects as spines, thoms, or small,bare stems on plants. Spines and thorns are presenton mesquite,hackberry tees, Celtis sp., skunkbush,Rftzs sp., pencil cholla(tasajillo),Opuntialeptocaulis DC var. leptocaulis,and silverleafnightshade, Solanum elaeagnifoliumCav., in the are4 but eggshave been located only on prickllpear spines. The antF. neoclarais commonin the studyarea, and many individualswere frequently seenclimbing on andunder the prickllpear cactus. Onelarge colony was locatedunder a fallen mesquitelimb nearthe prickllpear, but no C. dominicanafranciscanalarvaewere found in the ant colony. Erber (1988) indicatedthat the mynnecophilousClytrinae are carnivorous,but in oru rearingattempts, insect remains were rarely fed upon. Insectremnants, either freshor dried, werenotpaf,tofthedietin1995whenthetwolarvaewererearedtoadult.Erber(1988)reported

% tlntC. dominicanamaynotrequire an association with ants,andwhileAnomoeaflavoknwiensis Moldenke is associatedwith ants, it is not known if the antsare necessaryfor survival of the lanae (LeSageand Stiefel 1996).

ACKNOWLEDGMENT

This researchwas supportedby TexasAgricultural ExperimentStation Project H-8136. G. B. Idol andR. Montandon(Texas Agricultural ExperimentStation, Vemon) helpedlook for the eggclusters and assistedwith rearinglarvae, and K. Bamett preparedFig. I . I am grateful to E. G. Riley (collections Manager,Departnent of Entomology, Texas A&M University, CollegeStation) foride,lrtificationof Coleothorpa dominicanafranciseanaandforthe discussion and referencesregarding current taxonomic status,to J. L. Cook (Departmentof Biological Sciences,SamHouston StateUniversity,Huntsville, TX) foridentificationofFormicaneoclara, to S. L. Dowhower (Texas Agricultural Experiment Station, Vemon) for identification of Opuntia engelmanni,andtoE. Thonilson @epartne,ntofPlant andSoil Sciences,Texas Tech University, Lubbock) for information on Coscizoptera Wp.in the TexasTech University Insect Collection.

LITERATURE CITED

Erber,D. 1988. Biology of CamptosomataClytrinae - C4ptocephalinae- Chlamisinae- Lamprosomatinae,pp. 513 - 551. .InP. Jolivet,E. Petitpiene,and T. H. Hsiao [eds.], Biology of Chrysomelidae.Kluwer AcademicPublishers, DordrechV Boston/ London. Gilbert,A. J. 1981. A new speciesof CoscinopteraLacordaire from California(Coleoptera: Chrysomelidae).Pan-Pacific Entomol. 57t 364- 370. kSage, L., and V. L. Stiefel. 1996. Biolory of immaturestages of the North American clytrines Anomoealaticlavia (Forster) and,A. flavokansiensisMoldenke (Coleoptera: Chrysomelidae:Clytrinae), pp.2l7 - 238. In P. H. A. Jolivet andM. L. Cox [eds.], ChrysomelidaeBiology, vol. 3: General Studies. SPB Academic Publishing, Amsterdam,The Netherlands. Moldenke, A. R. 1970. A revision of the Clytrinae ofNorth America north of the Isthmusof Panama.Stanford University, Stanford. 310 pp. Moldenke, A. R. 1981. A genericreclassification of the New World Clylrinae (Coleoptera: Chrysomelidae)with a descriptionof new species.Entomologische Arbeiten ausdem MuseumG. Frey29:75- 116. Rilen C. Y. 1874. The Dominican Case-bearer- Coscinopteradominicana @abr.). Sixth annualreport of the stateentomologist (Missouri), pp. 127- l3l. Riley, C. V, 1882. Habitsof Coscinopteradominicana. Amer. Nat. 16: 598. Rilen C. V. 1891. Chrysomelidlarvae in ant'snests. Insect Life 4:148 - 149. shorey,H. H., andR. L. Hale. 1965. Mass-rearingof the larvaeof nine noctuidspecies on a simple artificial medium. J. Econ. Entomol. 58:522 - 524. Smith,L. L., andD. N. ueckert. 1974. Influenceof insectson mesquiteseed production. J. RangeManage.27:61 - 65. Stiefel,V.L.,andD.C.Margolies.1998. Ishostplantchoicebyacly{rineleafbeetlemediated throughinteractions with the antCrematogaster lineolata? Oecologia ll5: 434- 438.

% vol.28 NO.2 SOUTHWESTERNENTOMOLOGIST JuN.2003

HEAT SHOCKPROTEIN 70 DURING DEVELOPMENTOF THE LEAFCUTTING BEE.M EGACH I LE ROTUNDATA (HYMENOPTERA : MEGACHILIDAE)

JohnM. Hranitzrand John F. Barthell

Departmentof Biology,University of CentralOklahoma, Edmond,OK 73034

ABSTRACT

Expressionof a 70,000M. heatshock protein (HSp70) in Megachilerotundata (Fabr.) was studiedfrom diapauseto the pupalstage to determinewhether HSP70 levels change duringpupal development. we examinedlevels of HSp70inthe headcapsules of prepupie that were sampledfrom threetemperature regimes over l0 days:refrigerator (Rr; at o-z;c, room temperature(RT) at 2l-24"c, and incubator(IN) at 30"c. western blots usinga monoclonalantibody against HSP70 detected a singleband at 70 kDa in diapausingprepupae andduring pupaldevelopment. Enzyme-linked immunoabsorbent assays (ELISA) revealed only minor fluctuationsin meanlevels of HSp70during diapause and pupal development. overall, the percentageof HSP70in headcapsules of M rotundatarang"df.o- 0.g-1.7o/o. Although large variation in HSP70 levels among individual larvae was observed,no significant changesin HSP70 levels occurredduring post-diapausedevelopment to pupal stages.

.INTRODUCTION

The alfalfa leafcutting bee,Megachile rotundata (Fabr.), is an increasinglyimportant poflinator of seedalfalfa in theUSA (TorchioI 990,Peterson et al. 1992).It o-verwintersas a diapausingprepupa inside brood cells constructed within preexistingcavities. Unlike its native megachilid countetpartsin the northwestemUSA, M. rotundata will nest in open grasslandand, despite exceedingly high temperatures found there, construct nests in exposed habitatsthatcanexceed45oC(Barthelllgg2,Barthelletal.1998). Becauseoftheunusual lemperatureextremes encountered by this speciesand the numerousproteins differentially expressedduring development(Rank et al. 1982,Rank et al. t9g9), this beeprovides an opportunity to studyvariation in the levelsof HSps during its life cycle. Among the I I proteins with alteredexpression during pupal development of M. rotundata,the expression ofone 70 kDa proteinnotably declines in developingpupae and teneral adults (Rank et al. 1982). The patternof expressionof the 70 kDa proteinin M. rotundataappearssimilar to post-diapause pattemsofexpression for theHSp70 family Hspsinsarcopiiga crassipalpis (Joplinand Denlinger 1990) and Lymantria dispar (yocum et al. l99l). rPresent address:Department of Biologicaland Allied HealthSciences, Bloomsburg university of Pennsylvania,Bloomsburg, pA l7gl5. E-mail:[email protected] 97 The purposeof this study was to determinewhether M. rotundata poss€ssesa 70 kDa HSPthat undergoes altered expression during pupal development. We specificallychose to studyHSP70 expression in thehead capsule of M. rotundatabecause expression ofHSPT0 in the insect brain protectsnervous system function during development(Krebs and.Feder 1997)and because different tissues may expressdifferent isoforms in the HSP70family of HSPs (e.g.,Joplin and Denlinger1990). We usedSDS-PAGE and Westernblotting to determinewhether the 70 kDa bandin M. rotundatawas possibly a HSP70family HSP,and employedanenzymeJinked immunoabsorbent assay (ELISA)to quantifu HSP levels in head capsules.

MATERIALS AND METHODS

Diapausing M. rotundata prepupae(in cocoons) were obtained from International PollinationSystems (Manitob4 Canada).These were kept in a refrigeratorat 0-7oC to maintain diapause,a dormantstate with respect to developmentalmorphogenesis, prior to our experiment. We measuredthe amountsof solubleprotein and HSP70 family proteins in thehead capsule ofdiapausing and developing prepupae over a periodof l0 daysat three temperatureregimes: refrigerator (RF) at 0-7"C, room temperature(RT) at 2l-24"C, and incubator(IN) at 30'C. Theseconditions approximate those used during storageand incubationof M rotundatafor commercialpollination (see review in Petersonet al. I 992). Maximum and minimum temperatureswere recordedfor eachof the temperatureregimes, exceptthe 30'C (constant)regime, using a maximum-minimumthermometer (Taylor Instruments,Fletcher, NC). Becausepronounced expression of HSP70is well-documented in larval brain tissueof at leastone insect,Drosophila melanogasrer(Meigen), we choseto focuson changesin thehead capsule ofthe prepupae(Krebs and Feder I 997). Cocoonswere collectedfrom eachof thetemperature regimes on Days0, 2,4, 6, 8 andl0 of theexperiment and immediately frozen and storedat -80oC for subsequentanalyses. Teneral adults were collectedfrom the IN temperatureregime on Day 26. We identifiedthe molecularweight of HSP70family HSPsin prepupaeby SDS-PAGEand Westemblotting, and determined soluble protein and heat shockprotein concentrationsby spectrophotometry. Homogenateswere prepared using one (teneraladult) to five (prepupae)head capsules of M. rotundatahomogenized in cold l0 mM phosphatebuffered saline pH 7.6(PBS), with 0.2% sodiumazide and 2 mM tosyl argininemethyl ester (TAME). Homogenateswere centrifugedat 16,000x g for 20 minutesto remov€debris. Supematantsoluble protein was measuredby the Bradfordassay (Bradford 1976)and was the protein sourcefor SDS-PAGE and the ELISA. For Westemblots, we separated80 pg of solubleprotein using one- dimensionalSDS-PAGE as describedby Laemmli (1970). The gel consistedof a7%o acrylamideresolving gel (pH 8.6)and stackinggel (pH 6.8). Aliquotsofhomogenates were combined( I vol/2 vol) with samplereducing buffer containingbromophenol blue (0. I 7o)and denaturedfor 4 minutesat 95 oC. Proteinswere separatedby electrophoresisat 20 V for 8 hourson a minigel(Bio-Rad t aboratories,Hercules, CA) andelectrophoretically transfened to nitrocellulosemembranes overnight at 90 mAmpsin tris-glycine-methanolpH 8.6buffer usinga miniblot apparatus(Bio-Rad Laboratories, Hercules, CA). Generalproteins were detectedusing Ponceau-S stain (Sambrook et al. 1989),after which membraneswere washed l0 minuteswith distilledwater before immunodetection of HSPs.HSP70 family HSPswere immunologicallydetected using a monoclonalantibody (l :1000dilution) for bovineHSP70 (Sigma-Aldrich,St. Louis, MO) andanalkaline phosphatase-conjugated secondary antibody (Bio-RadLaboratories, Hercules, CA) at a l:3000 dilution. Threereplicate gels were run to evaluatethe consistencyofthe gel results.

q3 I

To quantifi HSP70proteins recovered from headcapsules ofM rotundata.wemodified the monoclonal-antibodyELISA describedby Yu et al. (1994). We coated Costar microplates(Bio-Rad Laboratories, Hercules, CA) with 80pg of solubleproteinperw€ll and allowed binding to occur overnight in coating buffer: 0.01 M sodium carbonateand bicarbonatebuffer (pH 9.6). We then washedthe plate with PBS-Tween(PBST, l0 mM PBS,0.05% Tween 20) onceand blocked wells for onehour at 37"C with PBST-BSA(l% BSA) to reducenonspecific binding of the antibodies.The plate was washedonce with PBSTand monoclonal antibodies for bovineHSP70 (Sigma-Aldrich, St. Louis,MO) were addedatal:1000dilution.Afteraonehourincubationat3T"C,theplateswerewashedfour times with PBST andthe alkalinephosphatase-conjugated secondary antibodies were added oC, at a I :3000dilution. Followinga one-hourincubation at 37 plateswere washed six times with PBSTand 200 rzl of thecolorimetric detection buffer were added to eachwell. After a 90 minute incubation at room temperature,the O.D.no,for eachwell was determinedby usinga DynatechMR5000 EIA reader(Dynex Technology, Chantilly, VA). Eachmicroplate containedHSP70 standardsconsisting of different dilutions of bovine HSP70(Sigma- Aldrich, St. Louis, MO). Threereplicates, each loaded in triplicateon microplates,were amlyzsd for each temperatureregime. We tested for differencesin soluble protein and HSP70concentrations among treatments on eachsampling day using the one-wayANOVA (Days 0-8) and t-test(Day l0) of SYSTAT 8.0. The experimentwiseerror rate for each variable was adjustedfor the numberof statisticaltests using the Bonferroni method.

RESULTS

Prepupaemaintained in the RF and RT temperatureregimes showed no obvious morphological changefor the l0 days of the experiment. In the IN temperatureregim€, prepupaeterminated diapause and experienced mirked morphologicaldevelopment by Day b whencocoons contained pupae. Immunodetection of HSP70on Westemblots of all M rotundatain diapauseor post-diapausedevelopment revealed one band at 70 kDa (Fig' I ).

t 1 2 3 4 f, 6 7 8 w{

&t*w*e '-r:n+ drd &e 66 kDA

FIG. 1. Westernblot of proteinswith mouseanti-bovine HSP70 antibodies. Each lane contained 80 pg of soluble protein from head capsulehomogenates of prepupae(5 per homogenate)or teneraladults (l per homogenate).Lanes l -3: prepupaefrom the RF, RT. and IN temperatureregimes (eft to right) on Day 0; Lanes4-6: prepupaefrom the R-F,RT' and IN temperatureregimes (left to right) on Day 8; Lane 7: teneraladult on Day 26;Lane 8: molecularweight markers.

I At Day 0, samplesof diapausingprepupae that were placed in each temperatureregime showednearly equal levels of expressionofthis 70 kDa heatshock protein band. Although equal amountsof solubleprotein were supplied in eachlane, amounts of HSp70on oay t appearedlower in RT and IN prepupaethan RF prepupae. Low amountsof HSp70, comparableto HSP70levels in RT andIN prepupae,were also observed in theteneral adults. No change in soluble protein concentrationof head capsulesoccurred among the prepupae from the three temperatureregimes (Table l). At the onset of the experiment, means for soluble protein concentrationin the RT and IN prepupaewere 13.4o/oandl6.So/o higher than RF prepupae,respectivery. By Day +, tre iotubte protein contentof head capsules of RT andIN prepupaedeclined to levelsthat were 8.5% and2}.5%lowerthan the RF temperatureprepupae. Solubleprotein concentrationin developingprepupae from the IN temperatureregime decrinedthrough Day g such that head"upri"r or iN p."pupu contained levels27.9 and 39.5% less soluble protein than their RT andRF counterparts.By Day 1 0, solubleprotein concentration ofhead capsulesofdiapausing prepupae ii trrelp ano RT regimes increasedfrom the leversdetected at the on.et orite exieriment. rupal development in the IN temperatureregime was acceleratedand mosi of the cocoons containedpupae on Day 10, with too few prepupaeto study HSp70 expressionin that temperalureregime.

TABLE l. SolubleProtein and HSp70concentrations in prepupaeDuring Diapauseand Post-DiapauseDevelopment.

lN"

(e.16) (14.23) (51.2e) (3.4s) (l.sl) (0.s1) 2 416.50 414.48 455.20NS 22.81 22.72 2t.29 NS (22.8s) (36.72) (30.41) (1.73) (2.84) (1.30) 4 562.51 5r4.62 447.34 NS 23.64 22.84 20.44NS (s4.re) (l10.23) (22.7s) (3.88) (3.2e) (0.57) 6 s68.74 462.tr 357.34NS 23.62 23.06 22.16NS (94.87) (8.e6) (r8.33) (1.80) (2.38) (r.38) 8 518.32 434.62 313.57NS 24.22 21.77 t9.67 NS (46.21) ( 15.2e) Q7.34) (r.e3) (2.s1) (r.08) l0 528.34 445.38 NA" NS 22.57 21.39 NA" NS (44.4r) (38.80) (3.72) (2.ee) uMeans(standard erron in parentheses)are for threereplicates per sampleday in each regime (RF = refrigerator,RT = room temperature,IN : incubaior). 'Prob"jelReratur.e = probabilityin a one-wayANOVA, NS = not significantat o = 0.05. T.lA = too few prepupaefor analysis.

Although _,- the highestlevels of HSp70occurred on Day 0, the meanconcentration of HSP70never differed among trea&nents. After Day0, meanHSP70 levels declined slightly, with small fluctuations through the remainder of the experiment(Table l). The same rankings 9cc-urydamong temperature regimes on all sampringdays, witi diapausing prepupae in the RF temperatureregime possessing the highesimean HSp70 concentrations and prepupae the IN containing the lowest mean HSP70concentrations. The variation in HSP70 concentrations,measured as the coefiicient of variation (cV) in HSp70 concentration,was greater in ll rotundata fromthe RF ( cV = l3.l%o- 2g.6o/oiand RT (cv tm = lO.5%- 24.9o/o)temperature regimes than in the IN (CV = 36% - 10.8%)temperature regime.The percentageof HSP70in soluble protein of homogenatesranged from 0.8% to l.7o/oanong all samPles.

DISCUSSION

We reporta 70 kDa HSPin headcapsules of M. rotundatathat is immunologicallyrelated to the HSP70family HSPs. This HSP70family proteinpossesses the samemolecular weightas a 70 kDa proteindescribed by Ranket al. ( 1982).Although HSP70 concentrations werehighly variableamong replicates, HSP70 levels accumulated to about0.8-1.7% ofthe soluble protein in the headcapsules of M. rotundata atthe temperaturesused to maintain diapauseand promote pupal development. Tbesepercentages are comparableto the percentageHSP70 reported in otherectotherms (e.g., Feder et al- I 997,Yu et al. I 994). Variation in levels of HSP70,which was two to three times higher in the diapausing prepupaethan in developingprepupae, may haveresulted from eitherontogenetic expressi on or individualvariation in stress-inducedexpression ofHSPTO in prepupae.An ontogenetic explanation for the variation HSP70 in M. rolundata is supportedby other observations showing that expressionof HSP70-familyHSPs during insect larval developmentcauses dramatic changesin the amount and number of isoforms presentin tissuesof different developmentalstages of insects(Joplin and Denlinger 1990,Krebs and Feder 1997). Similarly, a stress-inducedexplanation is plausible becausenumerous studies show that stressprotein expressionis sensitiveto both the magnitudeand duration ofa stressor(Feder and Hoffman 1999),a phenomenonillustrated by studiesreporting cold-shock induced expressionof HSPsin insects.A varietyof HSPsin insects,including those in the HSP70 family, arecommonly induced by cold-shock.Severe, acute cold-shock (-10 to -20"C for l-8 h) inducesHSP70 family HSPsin Sarcophagiacrassipalpis (Joplin et al. 1990,Joplin andDenlinger 1990) and Lynantria dispar (Yocum et al. l99l). Intensifiedseverity, either temperatureor duration,ofthe cold stressincreases the duration ofHSP expression(Joplin etal. l990,Yocumetal.l99l). EventhoughlowervariationinHSPT0levelscorresponded with rapid development at 30"C, suggesting developmental canalization of HSP70 expressionafter diapause, we cannot differentiate stress-induced expression ofHSPs among individualprepupae from developmentallycontrolledpattems ofexpressionduringdiapause and pupal development. While HSP70expression has been characterized in somespecies of Hymenoptera,earlier studiesshow that HSP expression in Hymenopteramay be induced bytwo differentstressors, heatshock and parasitic infections (Severson et al. I 990,Gehring and Wehner I 995, Gregorc and Bowen 1999). Although the functionalsignificance of HSPsduring diapauseand pupal developmentof M. rotundata remainsto be elucidated,others have described how high or low temperatureextremes can delay emergenceof M. ronndata from the pupal stage (Underragaand Stephen 1980a, 1980b). Small differences in temperature(< 5"C) canalso producesignificant differencesin survival amongpre-adult developmental stages as well as in the sizeand survivorship ofadult bees(Tepedino and Parker I 986,Richards et al. I 987, Whitfield andRichards 1992). Because heat-shock proteins can serye as useful biomarkers for environmentally-inducedphysiological stress (dePomerei 1996, Feder and Hofmann 1999), HSP70 levels may serveas a us€fiil bio-indicator of stressin developmentalstages of M. rotundata. lf HSP70expression in M. rotwrdataishighlysensitive to environmental stessors(e.g., toxicants, pesticides, stresses during storage or incubationofdevelopmental stages),then it could be usedtogether with traditional indices,such as emergencesuccess ratesand body size, to study storag€and rearingconditions for M. rotundata.

101 ACKNOWLEDGMENT

This researchwas supportedby Faculty DevelopmentGrants from the college of GraduateStudies and Research (University of centraloklahoma) to J. F. Barthelland.J. M. Hraritz. we greatlyappreciate R.w. Thorp(University of californi4 Davis).R. Bitner.D. Nahuliak,and S. Peterson(Intemational Pollination Systems) for assistanceduring the study aswell asW. Kempand G. Yocum(United States Department of Agriculture)for comments on the manuscript.we alzugratefully acknowtedge support from a grantprovided to S. N. Rao through the EPSCoRSummer Research Intemshi ip program and assistance in the laboratoryby D. Ngo andS. Khan.

LITERATURECITED

Barthelf,J.F. 1992. Heterogeneity,invaders, and the populationdynamics of someoak forest solitary bee communities. ph.D. Dissertation.University of Califomia, Berkeley. Barthell,J. F., G. w. Frankie,and R. w. Thorp. 1998.Invader effects in a communityof cavitynestingmegachilid bees(Hymenoptera: Megachilidae). Environ. Entomol.2T: 240-247. Bradford,M. M. 1976. A rapid and sensitivemethod for the quantitationof microgram quantitiesof proteinutilizing the principleof protein-dyebinding. Anal. Biochem. 72:248-254. Feder,M. E., and G. E. Hofmann. 1999. Heat-shockproteins, molecular chaperones, and the stressresponse: evolutionary and ecological physiology, Ann. Rev.physiol. 6l: 243-282. Feder'M. E.,N. Blair, andH. Figeuras.1997. Natural thermal stress and heat-shock protein expressionin Drosophilalarvae and pupae. Funct. Ecol. I l: 90-.|00. Gehring,W. J.,and R. Wehner.1995. Heat shock protein synthesis and thermotolerance in cataglyphis,an ant from the Saharadesert. proc. Natl. Acad.sci. USA 92:2994- 2998. Gregorc,A.. andl. D. Bowen. 1999 In situlocalization of heatshock and histone proteins in honey-bee(lp is melliferaL.)larvae infected with Puenibacilluslawae. Cell. Biol. lnt.23:211-218. Joplin,K. H., andD. L. Denlinger.1990. Developmentaland tissue-specific control of the heatshock-induced 70 kDa relatedproteins in the flesh fly. sar cophagacrussipatpis. J. InsectPhysiol. 36: 239-249. Joplin,K. H., G. D. Yocum,and D. L. Denlinger.| 990. cold shockelicits expression of heatshock proteins in the flesh fly, sorcophagacrassipalpis. J. Insectphysiol. 36: 825-834. Krebs,R. A., and M. E. Feder. 1997. Tissue-specificvariation in Hsp70expression and thermaldamage in Drosophilamelanogaster larvae. J. Exp. Bior. 200: 2ffi7-2015. L,aemmli,u. K. 1970. cleavageof structuralproteins during the assemblyof the headof bacteriophageT4. Nature(London) 227 : 680-685. Peterson,S. S,,C. R. Baird,and R. M. Bitner. 1992.Current status of thealfalfa leatbutting bee,Megachile rolundata, as a pollinatorofalfalfa seed. BeeSci. 2: 135-142. dePomerai, D. 1996.Heat shock proteins as biomarkers ofpollution. Hum.Envir. Toxicol. 15:279-285. Rank'G. H., A. J. Robertson,and S. M. Gilmer. | 982. An analysisof majorprotein species duringpupationin Megtchile roturulara.lnseot Biochem. Mol. Biol. 12:699-705.

la RanlqG. H., D. W. Goerzen,and J. H. Gerlach. 1989. Crossedimmunoelectrophoresis of major protein antig€nsduring pupationof the leafcuttingbee Megachile rotundata. Apidologie20: 139-147. Richards,K. W., G. H. Whitfield, and G. B. Schaalje.1987. Effectsof temperatureand dwation of winter storageon survival and period of emergencefor the alfalfa leafcutterbee (Hymenoptera: Megachilidae). J. Kans.Entomol. Soc. 60: 70-76. Sambrook,J., E. F. Fritsch,and J. Maniatis.1989. Molecular cloning: A laboratorymanual, 2d ed. Cold Spring Harbor, NY: Cold Spring Harbor LaboratoryPress. Severson,D. W., E. H. EricksonJr., J. L. Williamson,and J. M. Aiken. 1990. Heatstress induced enhancementof heat shock protein gene activity in the honey bee (Apis mellifera).Experientia 46 : 737-739. Tepedino.V. J.,and F. D. Parker.1986. Effect ofrearing temperature on mortality,second- generationemergence, and size of adult in Megachile rolundata (Hymenoptera: Megachilidae).J. Econ.Entomol. 70: 974'977. Torchio, P. F. 1990. Diversificationof pollinationstrategies for U.S. crops. Environ. Entomol. 19:.1649-1656. Undurraga,J. M., and W. P. Stephen.1980a. Effect of temperatureon developmentand survival in post-diapausingalfalfa leafcutting bee prepupaeand pupae(Megachile rotundata (F.): Hymenoptera:Megachilidae). I. High temperatures. J. Kans. Entomol.Soc. 53: 669476. UndurragaJ. M., and W. P. Stephen.1980b. Effect of temperatureon developmentand survivalin post-diapausingalfalfa leafcutting bee pupae (Megachile rotundata (F.)): Hymenoptera:Megachilidae). II. Low temperature.J. Kans.Entomol. Soc. 53:.677' 682. Whitfield, 0. H., and K. H. Richards. 1992. Temperature-dependentdevelopment and survival of immature stagesof the alfalfa leafcutter bee, Megachile rolundata (Hymenoptera:Megachilidae). Apidologie 23 : | | -23. Yocum,G. D., K. H. Joplin,and D. L. Denlinger. 1991. Expressionof heatshock proteins in responseto high andlow temperatureextremes in diapausingpharate larvae ofthe gypsymoth, Lymantria dispar. Arch.InsectBiochem. Physiol. 18:239'249. Yu,2., W. E. Magee,and J. R. Spotila. 1994. Monoclonalantibody ELISA testindicates that largeamounts of constitutiveHSP-70 are present in salamanders,turtle and fisb. J.Therm. Biol. l9:41-53.

1B vol.28 NO.2 SOUTHWESTERNENTOMOLOGIST JUN.2003

ATTRACTION OF DIAMONDBACK MOTH' TO TTNEB COMMERCIAL SEX PTIEROMONELURES UNDER LABORATORY AND FIELD CONDITIONS

Xaodun He, Wen Chen2and Tong-Xian Liu2

BeneficialInsects Research Unit, Kika de la Guza SubtropicalAgricultural Research Center, Agricultural ResearchService, U, S. Departmentof Agriculture, 2413 E. Highway83, Weslaco,TX 78596

ABSTRACT

Severalcommercial pheromone lures for diamondbackmoth, Plutella rylostella (L.) were usedfor monitoringmale mothsin the Lower Rio GrandeValley, Texas,in the 1990's;however, none were successfulin capturingmale moths. In this study,three commercialsex pheromonelures (Pheroconru, Scenturion ld Watchil, andIPMN) for P. rylosulla were tested in a cabbagefield and walk-in field cagesusing the scentry and rRECE traps. All three pheromonelures were effective for capturing adult males of P. rylostella under field conditions. Among the three lures, PheroconN and IpMru lureswere significantlymore effective in generalthan the Scenturionld Watch*. The effectivenessof the three lures lasted longer than seven weeks. There was no significant differenceamong the three lures in a no-choicetest in the field cages. Mating disruptionof the lureson maleP. rylostellawas observed in the two-lure-choice testsin the walk-in field cages.The ratio of the threecomponents, (Z)- I l -hexadecenal, (Z)-1l-hexadecenylacetate, and (Z)-ll-hexedecenolwas found to be 3:l:2 from l- to 6-d old P. rylostella female abdominaltips using GC-MS analysis. The level of pheromonesproduced by femalesreached a maximumfrom 34 d old females. The ratiosof (Z)- I l-hexadecenal,(Z)-1 I -hexadecenylacetate, and (Z)- I l-hexedecenolwere detectedas l.l5:1.00:0.04,1.60:1.00:Q.23 and 0.52:1.00:0.03 from the commercial pheromonelures of Pheroconm,Scenturion ld Watchruand IPMN, respectively.

INTRODUCTION

The diamondbackmotlq Plutella rylostella (L.), is one of the most destructive worldwide pests of crucifers such as cabbage,cauliflower, broccoli, watercress, mustard,and rape (Talekar and Shelton 1993). Texas is one of the largestU. S. producersof freshmarket cabbage with a yearly valueof $30-60million dependingon market prices (Anonymous2000). cabbage productionrepresents a $19,224,00Q industryin southrexas alone(Anonymous 2000). Averageinsecticide expenditure to

. LEPIDOPTERA:PLUTELLIDAE 2 VegetableIPM Laboratory,Texas A&M UniversityAgricultural Research and E:rlensionC€nter, 2415 E. Highway83, Weslaco,TX 78596

105 managelepidopterous and aphid pestsin southTexas is over $l million per year with no guaranteeof adequatecontrol due to the currentthreat of insecticideresistaice, especially with P. rylostella and cabbagelooper, Trichoplusia ni (Hiibner). After the mid-1980's,the diamondbackmoth becamea key pestin southTexas coincidental with the increasein use of pyrethroid insecticidesand subsequentdiamondback moth resistanceto them(Magaro and Edelson 1990). _ The sex pheromonesof P. rylostella consist of three compounds:(Z)-ll- hexadecenal,(Z)-ll-hexadecenyl acetate, and (Z)-11-hexadecenol.Lin et al. (i9g2, 1984)and Koshihara et al. (1980)reported that the mixtureof Zlt-16:Al, ztt-16:Ac. andzll'16:oH in the ratio of 5:5:0.1or 3:7:0.1were quiteattractive to malesof p. xylostellqin cabbage.Little informationis availableon the useof pheromonelures of P. xylostella in the Rio Grandevalley, Texas. The objective of this study was to comparethe effectivenessof three commercialsex pheromonelures (pheroconN, Scenturion 1"r watchfr, and IPIItrM) to attract male p. rylostella, to determinetheir longevity, and to comparethe chemical ratios of commercialpheromone lures with that producedby sexglands of femalesfrom the Lower Rio GrandeValley, Texas.

MATERIAL AND METHODS

Plutella rylostella larvaewere rearedat 25 t 2ac with a photoperiodof 14:10 (L:D). Larvaewere fed artificial diet (Sheltonet al. 1991),and adultswere fed a 5olo sugarsolution. Male adultsfrom I through5-d old were bioassayedsimultaneously to studyinsect response to pheromone.For pheromoneextraction, female adults aged I to 6-d wereused. All synthetic standardcompounds, (Z)-ll.hexadecenal (2ll-16:AJ), (Z)-ll- hexadecenylacetate (Z1l-16:Ac), and (Z)-ll-hexedecenol(Zll-16:OlI) were purchasedfrom BedoukianResearch, Inc. for GC-MS analysis. The standardswere usedto comparewith the commerciallurg products.The pheromonelures used in these experimentswere obtainedfrom TRECE Pherocon,Inc. (Grey septa,lot # 33600781 andlot # 3360331;Salinas, CA), IPM Technologies,Inc. (Redsepta, lot #A6239913; Portland,OR) and Scenturionl" Watch, Inc. (Grey septa,lot # 19711661;Ctinton, WA). The pheromonetraps were the Wing Kit K902 (Scentry,Inc., Billings, MT) and the lC trap (TRECEPherocon). Both trapsare the sametype usedcommercially for P. rylostella. Eachpheromone-impregnated rubber septum was suspendedwith an insect pin in the interiorcenter ofthe trap. Pheromone components were quantified by gas chromatograph-mass spectrometry(GC-MS) using a Hewlett Packard6890 GC coupled to a HP 5973 Network Mass SelectiveDetector with a Solgel-Waxcapillary column (SGE: 60 m x 0.25 mm ID x 0.25 pm film thickness).Carrier helium gas flow was I ml/min. Oven temperaturewas programmedfrom 60oto 250'C at 2}'Clmin. Maximum temperature was maintainedfor 10 minutes. The systemwas controlledby a HPMS Chemstation. Initial analysiswas performedin the electronimpact @I) mode(70 eV) with the mass spectrarange scanned from mJz40 to 550. Injectionwas performedin the splitless mode and the injector was purged after I minute. Injector and MS detectorwere maintainedat 250" and 280"C, respectively. Three injectionsof each samplewere made. The volatile emissionsfrom three different lures were analyzedby the methods of Mayer and Mitchell (1998) and replicatedthree times. One hundred pg of n- tetradecanewas usedas an internal standard. We recovered98.5Yo of the extraction standard.

106 Pheromoneglands of 1 to 6-d old females of P. rylostella were excisedin the morning. The eighth and ninth abdominal segmentsof P. ryIostella females *ere extendedby placingpressure on the abdomenusing forceps,then excised,extracted in methylenechloride (500 pl) for 10-20minutes, and groundand centrifugedat 11,600 rpm for l0 minutesat 4oC. Generally,20-40females were analyzedin eachage group. GC-MS methodswere used to quantiry the t}ree componentsfrom female sex glands. An internalstandard was usedfor the quantificationofthe compoundof 100 ng ofn- dodecane.We recovered98.0yo of the extractionstandard. llalk-in Field Cage Trial. Three polypropylene screened(5 threadsper cm) cages(6.5 m long x 1.8 m wide x 2.2 m high) supportedby a framewere placedin an alfalfafield. Cageswere placed l6m apart. Trapswere hung 0.3m abovethe groundon a woodenpole within the cages. Two tests,a no-choicelure test and a two-choicelure test, were conductedin the cages. In the no-choicelure test, one trap with one ofthe threelures was placedin the centerofeach ofthe thLreecages. In eachcage, 100 male P. rylostella adultswere releasedin late afternoonon the day of lure placement.The numberof maleP. rylostella caughtin eachtrap was countedat 24, 48 and72-h afrer release.There were three replicates per lure. In the two-choicelure test,two trapswith different lures were randomly placed in each cage about 6.2m apart, with moths releasedat a point midwaybetween the two traps. In eachcage, 100 male P. rylostella werereleased at 09:30. Capturedmale P. rylostella wererecorded at 1,2, 4,8, 12,and 24-h afterrelease. The two-choicelure tests were repeated three times. The cageswere randomlychosen for the lure combinations.These tests were conductedin November andDecember 2001 at the USDA-ARS,Vegetable IPM Laboratory,Weslaco, Texas. CabbageField Trial. This trial was conductedfor 8 weeksfrom 25 September to 20 November2001 at the TexasA&M AgriculturalResearch and Extension Center, Weslaco,Texas. Cabbagewas plantedon 4 September2001 on a l-m bed with 30-cm spacing. The field sizewas 92.Im long by 3.lm wide. Pheromonetraps were hung in the cabbagefield on woodenstakes 10.2m apart from 0.2 to 0.3m abovethe ground. One of eachlure type was replicatedtkee timesfor a total of nine trapsin the test plus threecontrols without lures. Eachtrap was movedto the next trap's positionto avoid locationbias after eachweekly examination. During the testperiod, temperature ranged from 1lo to 36oCand the relativehumidity from 1l to l00yo. Naturallyoccurring male diamondbackmoths were present and these were relied on for evaluatingthe lures. The total number of P. rylostella captured per week per lure was used for statisticalanalysis for the cabbagefield data. The capturerates were calculatedfor no- and two-choice pheromonelure tests. The capture rate (Yo) is the percentageof capturedmoths over total releasedmoths. For the cabbagefield test dat4 the analysis includedlure and week as main effectsand the interactionbetween them. The no- choice-luretest data was analyzed,using a one-wayclassified model. The cageand lure nestedwithin the cagewere factorsused for the two-choicelure test. All testswere performedusing SASPackages (SAS Institute 2000).

RESI.JLTSAND DISCUSSION

Walk-in Field Cage Trial. In the no-choicelure tests, numbersof male P. rylostella captureddecreased with increasingtime following malerelease (Fig. l). The majority of P. rylostella were caught in the first 24-h. The capture rates of male P. rylostella on Day I were35Yo, 43Yo, and 5*/o from the ls Watihru, pheroconru,and IPM'- lures, respectively. On Days 2 and 3, the capturerate of male P. rylostella caughtwas lessthan 4Yo. Capturerates were not significantlydifferent among the three

TU 80

70

60

q) + s 50 c) € 40 c) = 30 F 20

l0

0 2 Days FIG' 1. Capturerates of maleP. rylostella capturedin the no-choicelures test in walk- in field cages(white bar: l"t Watchru,black bar: pheroconrM,diamond bar: IpMrM).

60

{n

[;40 n E

&€ro

Ero

9:30-10:3010:30-ll:30 1l:30-13:30 l3:30-17:30 t7:3O-2t:3O 2l:3O-9:30 Total24Hr

Tirne

FlG. 2a. capture rates of male P. Dtlostella captured in the two-lure-choice tests: nrIPru lwhitebar) versus Pheroconru @lack bar) in walk-in field cages.

IG lureson Days 1, 2, and3 or overthe 3-d periodof observation. In the two-choice lure tests, capturerates of P. rylo*ella were less than3Yo on the secondday for all lules. The majorityof maleadults were alsocaptured within the first day. The Pheroconru and IPMfM lures attracted 33% (P = 0.000b and 23% (P = 0.0038),respectively, more maleP. rylostella(Figs. 2b and2c)than li Watchru lures. Theseresults showed a similarpattern to that of the no-choicelure tests(Fig. l). When Pheroconw and IPMru lures were placed in the same cage, they had a lower total capturerate @ig. 2a) than the other combinationsof the lures !n other cages(Figs. 2b and 2c). The differencein total capturerates between PheroconrM and IPMrM lures was about7Yo (P = 0.2381). Reductionof the capturerate with thesetwo luresmay have beencaused by the amountofthe pheromonesreleased, affecting the ability ofmales to follow an odor plumeto its source.This is known asthe matingdisruption effect (Chow 1990) and has been used as a control techniqueto prevent males from locating individualfemales for mating. The capturerate alsocould be affectedby temperature, wind direction,and wind speedin the field. Reddyand Urs (1996) reportedthat the peak period of attraction for male diamondbackmoth was from sunsetto midnight; however,there was no obvious peak for the capturerates within the first 24-h period in this study. This observationwas similarto that of Furlonget al. (1995)who reported that maleP. rylostella were attractedto syntheticlure throughoutthe initial 24-h period. CabbageField Trial. Numbers of male P. rylostella attracted varied greatly amongthe threecommercial lures (Fig. 3), whereasthe no-lurecontrol traps caught no male P. xylostella. Both week and lure effects on the number of male P. rylostella capturedwere significantlydifferent (P = 0.001),but the interactionbetween these two main effectswas not significant(P : 0.9363)(Table 1). Attractionwas highest(28-55 moths)during Week 4 for all lures. Therewere no significantdifferences (P = 0.1641) betweenattractiveness of IPMru and Pheroconrulures. whereas the capturenumbers of the ld Watchrt lure *ere significantlylower thanIPMru (P = 0.0la7i andPheroconru lures (P : 0.0002)(Table 2). The rubbersepta lures effecfively captured male moths for morethan 7 weeks,similar to the findingsof Reddyand Urs (1996). Underthe field conditionsin the Rio GrandeVatley, Texas, we recommendthat luresbe replacedevery 7 weeksfor maximumeffrciency in attractionof maleP. rylostella.

TABLE 1. Analysisof VarianceParameters for Attractionof P. rylostella Malesin the C"UU"g.fUd trUt SourceofVariance df F-value P-value Model 23 Week 7 3.83 0.0013 Lure 2 7.94 0.0009 Weekx Lure t4 0.50 0.9363 Error 48 CorrectedTotal 7l

TABLE 2. contrastofNumber of P. rylostellaMales captured among Different Lures for the CabbageField Trials

Contrast Difference t-value IPMru vs Pheroconru -4.8 -1.41 0.1641 IPMrMvs lst WatchrM 8,7 2.52 0.0147 Pheroconruvs lst WatchrM 13.6 3.93 0.0002

1D H40 n

vs €so I

uzv

9:30-10:3010:30-11:30 ll:30-13:30 t3:30-17:30 tT:30-21:30 2l:30-9:30 Total24Hr Time FIG. 2b. capture rates of male P. rylostella captured in the twolure-choice tests: IPM'' (black bar) versus lo Watchru (white bar) in walk-in field cages.

60

50

4H40 + >s v €ro & o

,q )i

9:30-10:3010:30-ll:30 ll:3G13:30 13:30-17:30 17:30-21:30 2l:30-9:30 Total 24 I{r Time Fig. 2c. Capture rates of male P. xylostella caDturedin the two-lure-choicetests: PheroconrM(black bar) versusld WatchrM(white bar) in walk-in field cages.

ll0 tI] v) o+l ti !o 1i >ids) 9

We€k FIG. 3. Numbersof male P. rylostella capturedusing three commercialDheromone lures(black bar: IPMTM,white bar: l't watcirM, diamoid bar: pheroconrt;'in cabbage field (Weslaco,TX).

chemical differencesamong the commercialpheromone lures were analyze{ify Gc-Ms. The ratios of pheromonecomponents. varied among the three commeiciil lures (Ta9l.e3). TLe.ratio of the ld watchrM lure was d-=ifferentfrom those of Pherocon'-and IPM'- lures. Theratios of Zl l-16:4l, Zll-16:Lc, andZl l-16:OHare ilPortant to the capturerate ofP. rylostella (Schroederet al. 2000). The emitted ratios of Zll-16:4l to Zll-16:Ac were0.52-1.60:1.0 from threedifferent lures, which was within the rangeof 0.6- 4.0:1.0Qlayer andMitchell 1999)using septa. At week 8 of the test, the luresfrom PheroconrMand IpMrM still attracied*i'le udultsalthough the Zl l-16:oH was not detectable,indicating that the 2ll-16:Al andZl l-16:Ac are major attractive componentsto male adults. Howwer, thesepheromones are not very staLle becauseof the Z-doublebond isomers. we also detectedother isomers from the lures, suchas Ell-16:oH, andEll-16:Ac. Thesefindings emphasize the needfor rigorous quality control in the preparationand testing of pheromoneformulations. The inefficiency of the lures could causedecrease in maiing disruptionand reductionin captureof the male moths@oelofs et d. 1979). FemaleP. rylostella producethree pheromonecomponents: zrl-r6:N, zI- 16:Ac,and zll-16:oH from sexglands (Tamaxi et al. 1977,Koshihara and yamada l98l). The amountof pheromoneswas the highestfor the 3 to 4-d old femaleadults a). The Gig. ratio of Zll-16:4l, Zll-16:Ac, andZll-16:OH was foundto be 3:l:2 , wlri9h was in agreementwith the resultsof chow et al. (lgzs). They reportedatraction of virgin females increasedwith age, reachinga maximum attraction on days 4 and 5. However,Reddy (1996) and urs reportedthat five-day-oldvirgin - femalei rylostella did not attractmale moths because they lackedsex pheromone.

111 TABLE 3. Ratiosof (Z)-1l-Hexadecenal,(Z)-1l-Hexadecenyl Acetate and (Z)-11- Hexedecenolfrom ThreeCommercial P. rylostella SexPheromone Lures before and ifter EightWeeks of the CabbageField Trials UsincGC-MS Analysis Zll-16:N Zll-16'.Ac Z7l-16:OH Before It watchru 1.60 1.00 0.23 Pheroconru Ll5 1.00 0.04 IPMTM 0.52 1.00 0.03 After 1owatchrM 0.68 1.00 0.06 Pheroconru 0.36 1.00 ND PMTM 0.23 1.00 ND ND: NotDetectable

o, F

Ageof Insects@ays) FIG. 4. Effects of age of femaleP. rylostella on Pheromones(black bar: Zll-16:Ac, white bar: 211-16:OH,diamond bu: Zll-16:N) in CI{zCbextracts of abdominaltips usingGC-MS analysis.

Our resultsshowed that all three commercialpheromone lures placed in Scentry or TRECE traps effectively attracted male P. rylostella. Sex pheromonebaited trap techniquescan be useful in an integratedpest management system for insectdetection and surveillance,as well as for the timing, evaluation,and control ofpopulation levels of diamondbackmoth in the No GrandeValley, Texas. This techniquecan be usedto monitor P. rylostella population,which will aid in the decisionmaking processfor manaSementof this importantpest.

TI2 ACKNOWLEDGMENT

We -acknowledge the technical assistanceof Renee Sauer (JSDA_ARS, weslaco,TX) and HaseebMuhammad (Texas A&M university AREC, *estaco, TX). "]tq Y^. thank Aijun Znang (USDA-ARS, Beltsville, MD), atton N. Sparks(Texas A&M- university, AREC, weslaco, TX) and walker A. Jonis (usDA-Ais, wesraco, TX) for their commentsand suggestioni,and Lizhen wang (Prc usA ;t-krin, Ky) for the statisticalanalyses.

LITERATURE CITED Anonymous. 2000. Texas Annual Vegetable Summary. Website: http://www.io.com/-tass/tvegansm.htm. choq s., c. L. Hsu, y. ! and u.Lin. 197g. Field attractionexperiment of the sex pheromoneof the pruteila diamondbackmotlr" ryrosteila G..j in iaiwan. r.rat. Sci.Coun. Month. 6: 651-656. chou Y, s. 1990. Disruptioneffect of the syntheticsex pheromone and its anatogues on diamondback moth. pp. 105-108./'r N.S. Tarelar tedl DiamondLck Moth Management, proc. Second International workshop,' ro-r+ oec. 1990. A\IRDC, Taichung,Taiwan. Furlong,M. pell, p. J.' J. K. o. choo, and s.A. Rahman. 1995. Fierd and laboratory evaluation of a sex pheromonetrap for the autodisseminationoi the fi.rngal entomopathog en,zoophthora radicans @ntomophthorales) by the diamondback yponometidae). yt9lhtllutella rylostella (Lepidoptera: nuil.'nntomol. Res.g5: 331-337. Koshihar4 T., and H. Yamada. 19g0. Attractant activity of the female sex-Juf. pheromone of diamondback ptuteila _rgrl rytostelta (L.) and anatog*. l, Appf. Entomol.Zool. 24: 6-12. Koshihara, yamada. T., and H. r9gl. Femalesex pheromoneof the diamondback __ry9ttt, Plutella rylostella. Jan.Agric. Res.euarierly 15: 22_28. Lin, Y' M'r Y-' s. chow, and-H.c. TzJng. 1982. Field irapping of the diamondback moth Plutella xyroflel/a pseudarefia (Linnaues) and tiloin using the synthetio ldater; sexpheromone of the diamondbackmoth. s;uil. rnrt. zoor.,'ecaoemia Sinica2l:l2t-127. Lin, Y. M., Y. S. Chow, H. C. Tzeng,C. B. Hurng, and S. S. Wang. 19g4. The distanceeffects of(Z)-9-tetadeienyr acetateand (z)-rr-hexadeJen-r-oron the sexual communication of diamondbackmotta pruteira tytort a" u. Buil. Inc. Zool., AcademiaSinica 23: lg7-192. Magaro,J. J., and J. v. Edelson. 1990. Diamondbackmoth (Lepidoptera:ptuteflidae) in south Texas: A techniquefor resistancemonitori.g ii, irti i*ra J. Econ. Entomol.83 : l20l -1206. Mayer, M. S., and E. R. Mtchelr. 199g. Rapid meazureof sex pheromoneemission from ptastic- rope..dispense.r'""u'npi" of utirity i" ,;;i-Fivrop**i ;;;nication disruption of the diamondbackmotrr, ptuteila tyiostetta. rr7-r25. r* zo, Mayer,M. s., E. R. Mitcheil. -1999. Differencesbetween diamondback moth, pruteila (Lelidontera: pruteflidae), ylostella .-(L ) sex pheromoneslures are not determinable throughanarysis of emission.'Agric.Fbr. Entomol. t, iis-zle.

113 Reddy, G. V. P., and K. C. D. Urs. 1996. Studieson the sex pheromoneof the diamondbackmoth Plutella rylostella (Lepidoptera:Yponomeutidae) in India. Bull. Entomol.Res. 86: 585-590. Roelofs,W. L., T. W. Brooks, T. W. Burkholder,R. T. Carde,D. L. Chambers,H. H. Shorey,D. L. Wood, G. F. Ludvik, and P. L. Russell. 1979. Establishing efficacy of sex attractantsand disruptantsfor insect control. Entomological Societyof Americ4 CollegePark, MD. SAS Institute. 2000. SAS/STAT guide for personalcomputers, version 7. SAS Institute,Cary, NC. Schroeder,P. C., A. M. Shelton,C. S. Fergusor\M. P. Hoffmann,and C. H. Petzoldt. 2000.Application of syntheticsex pheromone for managementof diamondback moth,Plutella rylostella, in cabbage.Entomol. Exp. Appl. 94:243-248. Shelton,A. M., R. J. Cooley,M. K. Kroening,W. T. Wilsey, and S. D. Eigenbrode. 1991.Comparative analysis of two rearingprocedures for diamondbackmoth (Lepidoptera:Plutellidae). J. Entomal.Soc.20: 77-26. Tamaxi,Y., K. Kawasaki,H. Yamada,T. Koshihara,N. Osaki,T. Ando, S. Yoshida, andH. Kakinohana.1977. (Z)-l l-Heexadecenal& (Z)-l l-hexadecenylacetate: sex-pheromone components of the diamondback moth (Lepidoptera: Plutellidae).Appl. Ent. Zool. 12:208-10. Talekar,N. S. and A. M. Shelton. 1993. Biology, ecology,and managementof the diamondbackmoth. Annu.Rev. Entomol. 38:275-301.

tr4 vol.28 NO.2 SOUTHWESTERNENTOMOLOGIST JL'N.2003

PHEROMONEPRODUCTION IN CORNEARWORMI : EFFECTOF TEMPERATUREAND HUMIDITY

AshokK. Raina

FormosanSubtenanean Temrite ResearchUnit, USDA, ARS., ll00 RobertE. LeeBlvd., New Orleans,LA70l79

ABSTRACT

Femalesof thecorn earwonn,Helicoverpa zea (Boddie), produce a sexpheromone to attract conspecificmales for mating. Laboratoryexperiments were conductedto studythe effect of threetemperatures and three levels ofrelative humidity (RH) on pheromonetiter in the females measuredas the quantityof Zll-hexadecenal,the major componentof corn earwonnsex pheromone.Under normal ternperature and RH condition,females produced an average of 36.I + 3.9ng. lowest pheromonetiter (2.8+ l.l nglfemale)was observed in femaleskept at cool (l5oc) temperatureand normal(50%) RH. warm temperature(35"c) with any of the RH combinations(10, 50, or 9OYo)resulted in low pheromoneproduction. Pheromone titer could beincreasedsigrificantlybyinjectingpheromonebiosynthesis-activatingneuropeptide@BAN) into females maintained at cool temperature;this treatnent had no effect at the warm temperature.Thus,low availabilityofPBAN seemsto be at leastone ofthe factorscontributing to the decreasedpheromone titers at the cool temperahrre.

INTRODUCTION

Female-producedsex pheromones are cornmonly utilized by malesfor matefinding among lepidopteranspecies. Sex pheromone of the corn earworm, Helicwerpazea (Boddie)is a blend of four components,of whichZll-hexadecenal accountsfor approximately92Yo (Klun et al. 1980). Earlier studies referred to relative percentagesrather than absolute amounts of pheromonalcomponents. The first quantitativedetermination ofpheromone titers from single com earworrnfemales was reported by Rainaand Klun (1984). Pheromoneproduction is regulatedby endogenousand exogenousfactors (Card6 and Webster1980). McNeil (1991)reviewed the role of factorsaffecting emission and reception of pheromones in moths. Among the exogenousfactors, the roles of photoperiod and temperaturewere studied extensively. Almost all thestudies involving these factors measured the incidenceof calling without actual determinationof pheromonetiters. Thus calling, a behaviorassociated with releaseof pheromone,was reduced sigrificantly in the codling motfi, LaspeyresiapomonellaL,,when temperature decreased from 23" to l6"C (Castrovilloand Card6 1979). Calling was eliminated completely in the Oriental fruit moth, Grapholita molesta @usck) at temperatureswanner than 32oand cooler than 15"c (Baker and card6 1979\. However,very little informationis publishedon the effect of relative humidity (RH) on ' Helicoverpa zea @oddie) (Lepidoptera: Noctuidae) 115 pheromone production/calling(McNeil 1991). Baker and card6 (1979)reported that RH ranging from 25 to 100%did not affect calling in the Orientalfruit mottr. Studiesof the European com borcr,Ostrinia nubilatis (Hibner), revealed that the onset of callingwasearlier andthe proportion of femalescalling greaterat higherhumidities (Webster and Card6 19g2, Royerand McNeil 1991). The pheromone biosynthesis-activatingneuropeptide (PBAN) regulates pheromone productionin several speciesof moths(Raina I 993j. Rainaet al. ( I 99I isuggestedthat low pheromone productionin com earwormfemales at l4oCmightbe airetoinfrilUition of pBAN releaseand lower pheromonebiosynthesis. This study was conductedto determinethe effect of temperatureand RH on pheromone titer in com earwormfemales and to examinethe role of pBAN in regulatingpheromone productionunder experimental conditions.

MATERIALS ANDMETHODS

Weekly shipmentsof eggsof corn earworrnwere received from the InsectBiology and Population ManagementResearch Laboratory, Tifton, GA. ln Beltsville,MD, the "ggi*"re allowed to hatchand the larvaeplaced on artificial diet (SouthlandProducts, Lake Village, AR) in environmentalchambersmaintained at 26:21oc,60+ 5% RH, anda r6:d @:D)photoperiod with lights off at 0800and on at 1600hours EDT. After pupating,males and iemalis werekept in separateenvironmental chambers, Adults were provided l0% sucroseas food. Pheromone - titers(as quantity of Z1l-hexadecenal) were determined from femalesat 3-4 hours into the third scotophaseand quantified by gas chromatographyusing the intemal standardmethodas describedbyRainaandKemp eQ9t4. Comuinaf,onsofthreetemperatures andthree relative humidities were usedto determinethe effect on pheromoneproduction. For temperatures,15o,25o, and 35oc were consideredas cool, normul,and warm, respectively. similarly, 50% RH was consideredas normaland l0 and 90% RH as low and rrilrr. Higrr humidity (90%) was obtainedby placing a cage with adults inside a larger cage,and then sprayingwater into the outer cagewhich subsequentlywas covered with a p-olyethylenebag. A hygro-thermograph with a remote sensorplaced inside the cagewith ttre anuttswas usedto continuously monitortemperature and RH. Six newlyemerged females were placed under each test condition,and the pheromone titer was determinedafter 5 t hours(three houn into the third scotophase). To determine whether the low pheromoneobserved under some of the experimental conditionswas caused by lack of PBAN production/release,females under each experimental condition were dividedinto two groups,each consisting of ten individuals. ono groupwas injectedwith 20 pl pBAN salineand the second group with 5 pmol @eninsula,Belmont, cA) pl in 20 saline at one hour into the third scotophase.Pheromone was extractedfrom excised ovipositorsthree hours after injection ofthe peptideand quantified using GC analysis. For the experiments pRocMxED with temperatureand RH, lsAS kntitute 1996)with temperatureand RH as the effects was usedto analyzethe pheromonedata as a two-factor generallinear model' Data on the effect of PBAN were anallzed as a three. factor general linearmodel, with temperature,RH, and treatnent (checkorpBAN) as the effects. To corect fo-r varianceheterogeneity, treatments were groupedinto similar variancegroups for analysis. Meanswere compared using pair-wise contrasts.

RESULTSAND DISCUSSION

Pheromonetiters under various combinationsof temperatureand RH are shownin Fig. l. Pheromoneproduction was generallylow (36.1 + 3.9 ng/female)even at the optimal

r16 temperatureand RH conditions. Least pheromoneproduction (2.8 + l.l ngifernale) was observedat cool temperatureand normal RH. Comparingthe effectsof temperatureand RH on pheromoneproduction, statistical analysis ofthe dataindicated that the effectoftemperiture andthe interactionbetween temperature and RH werehighly significant(F :36.4, df :2, P < 0.0001and .F.: 54.7,df -- 4, P < 0.0001,respectively). At cooltemperahres, changing the RH to either low or high resultedin a significantincrease in pheromoneproduction (P < 0.05). At high temperahueand any combinationof RH, little pheromone was produced.At normal ternperature,both low andhigh RH causeda significantreduction in pheromonetiter (P s 0.05).

fi3s U) :30 o z2s o E,o Frs N ot c10

FIG. 1.Sex pheromone titers of laboratory-colonycom earwormfemales kept at combinations of threetemperatures (cool = I 5o,normal = 25o,and warm : 35"C)and three RHs (low = 10, normal : 50 andhigh :90%). Pheromonewas extracted34 hours into the third scotophase. Average+ SEM, N:6

Injection of PBAN had no significanteffect (P < 0.05) on pheromonetiter at high ternperaturein combinationwith low, normal, or high RH (Fig. 2). The maximal effect of PBAN injection was evidentat low temperaturein combinationwith normal andhigh RH (P< 0.001).Atnormal temperature,PBAN injectionresultedin a significantincrease ofpheromone titer only at normal andhigh RH. In nature,pheromones play a key role in matelocation arnongnoctumal species of moths. Phetomoneproduction and release are separate events and, whereas, we know muchabout how pheromoneproduction is regulatedin manyspecies of moths,not muchis known aboutthe intrinsic factorsthat regulatethe releaseof pheromone.Among the extrinsic factors,the role of light, particularlymoonlight, on the performanceof light hapshas been studied extensively (Hardwickl972,Mizutari 1984,Dent andPawar 1988, Yela andHolyoak 1997). Similarly, the effectsoftemperature and other factorshave been studied and reviewedthoroughlyby McNeil (1991). In almostall of thesestudies, calling was usedas a measureof pheromonl release,Pheromone titer asa measureofpheromone production has rarely been used. Exposure

Lt7 of corn earworm femalesto the warmer tempsrature(35"c) in the presentstudy resultedin significantly low pheromone titers. varying the humidity from l0 to 90% at this temperature hadno effecton pheromone titer. Giebultowicz et ai.(le92)reported thartemperatures df 33o and35"c causedpheromone titer to decreaseto amost o intatoratory-reareo-La wild gypsy moth,Lymantria dispar (L.) females,respectively. ono (1994)reporteo trrat iemates of the tubermoth produced n9!t9 very low amountsof pheromoneat 15"c for the first 2 daysof adultlife. In pseudaletia the true annywonn, unipuncta(Haw,), Delisle andMcNeil (19g7) reportedthat lowering the ternperaturefrom 25oto i oC causeoa sirlt in cattingbut no decrease in pheromonetiter.

r=so o +l F40 o 3 s30 g

:20 N Pro

NTNH NT/HH Exp€rimentalConditions

FIG. 2. Pheromonetiter of laboratory-colonyfemales kept under nine combinations of temperatures and RH. Check femaleswere injectedwith 20 pl salineand test femaleswith 5 pmol PBAN in 20 pl saline.Asterisks indicate significant difference (* p<.005,** p<0.001) betweencheck and test under each experimental condition. N = 10.

The least pheromone titer was found in fernales maintained at 5O%oRH and cool temperature.Exposure to high or low RH at the cool temperaturesresulted in a sigrificant increasein pheromonetiter. Apparently,the two extremehumidities partially suppressedthe eJfectof cool temperature.In all of thesecases, the pheromone titer was aboui 10ng or greater thanthatwhichmaybethelowthresholdforcalling.Rainaatel.(1991)observedarelationship betweeninitiation of callingand pheromone titer in com earworm.Pheromone titer in females at the time of initiation of calling during the scotophasewasl0 ng or greater. In Beltsville during the corn-growing season(June - September)of 1997,the average weeklylow temperatureconsistentlywas about l2oC except during2 weeks when itwas >1506. similarly, the averagedayime high for the sameperiod did not exceed350c, except dwing this same2-week interval (MD StateClimatologist Office). Temperatureswere coolest late at night or in earlymoming; whereas,most of the sexualactivity by corn earwormadults is confinedto dusk and the early part of the night (Agee 1969).Insects probably will not consistently encountertemperatures of 350Cor warmerduring evening/earlynight. During a periodof cool temperaturescombined with high humidity afteran eveningshower in summer,females should be ableto producesufficient pheromone (Fig. 1) to effectivelyathact males. Agee (1969) statedthat ideal conditions for reproductiveactivity includedRH between50 and 100%and temperaturesfrom 2lo to 280C.

118 PBAN had no sigrificant effect in elevatingpheromone levels at warm temperature, indicatingthat low pheromonetiter may havebeen due to low pheromonebiosynthesis and not lack of PBAN, low pheromonetitsr at warm temperaturealso mayhave resulted liom ahigh turnover (high synthesiscombined with high release/degradation).However, at.cool temperature,PBAN had a significant effect, leading to increasedpheromone production, suggestingthat releaseof PBAN was the limiting factor at cooler temperatures.Insects generally are inactive at cooler temperatures.Therefore, even if females were to produce pheromone,the males might not be ableto fly to thecalling female. Elsey (1982) reported that matingin thepickleworm, Di aphanianitidaft's (Stoll), occurredwhen temperature was wanner than l6'C. Sigrificant increasein pheromoneproduction caused by injection of PBAN into fernaleskept at normal temperatureand normal humidity was expected.

ACKNOWLEDGMENT

I thank Patricia T. Laernontfor meticuloustechnical assistance,Christopher Florane for making the figures and Mary J. Camp for statisticalanalysis of the data.Thanks to Jerome Klun, Ring Card€,and SonnyRamaswamy for critically reviewing an earlier version of the manuscript,This article reportsthe resultsof researchonly. Mention of a proprietaryproduct doesnot constitutean endorsementor a recommendationbv USDA for its use.

LITERATURE CITED

Agee,H. R. 1969. Mating behaviorof bollwormmoths. Ann. Entomol.Soc. Am. 62: ll20- 1122. Baker,T. C., andR. T. Card6.1979. Endogenous and exogenous factors affecting periodicities of female calling and male sex pheromoneresponse in Grapholitha molesta@usck). J. lnsectPhysiol. 25: 943-950. Card€,R. T., andR. P. Webster.1980. Endogenousand exogenous factors controlling insect sexpheromone production and responsiveness, particularly among Lepidopter4pp.9TT- 991..I2Regulationofinsectdevelopmentandbehaviour.Inst. Organ. Phys. Chem, Wroclaw Tech.Univ. Sci.Pap. No.22 (2). Castrovillo,P. J.,and R. T. Card€.1979. Environmental regulation of femalecalling and male pheromoneresponse periodicities in the codlingmoth (Laspeyresia pomonella). J. Insect Physiol.25:659-667. Delisle, J., and J. N. McNeil. 1987. Calling behaviourand pheromonetitre of the true annywonn Pseudaletia unipuncta (Haw.) (Lepidoptera: Noctuidae) under different temperatureand photoperiodic conditions. J. InsectPhysiol. 33: 315-324. Dent, D, R., and C. S. Pawar.1988. The influenceof moonlightand weather on catchesof Helicoverpa armigera (H$bner) (Lepidoptera:Noctuidae) in light and pheromonetraps. Bull. Entomol.Res. 78: 365-377. Elsey,K. D. 1982. Photoperiodand temperature effects on the occurrenceand periodicity of mating in pickleworm moths(Lrpidoptera: Pyralidae).Florida Entomol.65: 466471. Giebultowicz,J. M., R. E. Webb, A. K. Raina, and R. L. Ridgway. 1992. Effects of temperatureand age on daily changesin pheromonetiter in laboratory-rearedand wild gypsymoth (Irpidoptera: Lymantriidae).Environ. Entomol. 2l : 822-826. Hardwick,D. F. 1972.The influence of temperatureand moon phase on theactivity of noctuid moths.Can. Entomol. 104: 1767 -1770. Klur; J. A., J. R. Plimmer,B. A. Bierl-Leonhardt,A. N. Sparks,M. Primiani,O. L. Chapman, G. H. Lee,and G. Lepone.1980. Sex pheromone chemistryof female corn earworm moth, Heliothiszea. J. Chem.Ecol. 6: 165-175.

119 McNeil, J.N. l99l . Behavioralecology of pheromone-mediatedcommunication in mothsand its importancein theuse of pheromonetaps, Annu.Rev. Entomol. 36::407-430. Mizutani, M. 1984. The influenceofweather and moonlight on the light trap catchesof nioths. Appl.Entomol. ZooL 19: I 33-141, Ono, T. 1994.Effect of temperatureon biosynthesisof sex pheromonecomponents in potato tuberwormmoth, Phthorimaes operculella (Lepidoptera: Gelichiidae). J. Chem.Ecol. 201 2733-274t. Raina,A. K. 1993.Neuroendocrine contol of sexpheromone biosynthesis in Lepidoptera, Aruru.Rev. Entomol. 38:-329-349. Raina,A. K., andJ. A. Klun. 1984. Brain factorcontrol of sexpheromone production in the femalecom earwonnmoth. Science225l.531-533. Raina,A. K., andT. G. Kempe. 1992. Structureactivity studies of PBAN of.Flelicoverpazea (Lepidoptera:Noctuidae). lnsect Biochem. Mol. Biol. 22:221-225. Raina,A. K., J. C. Davis, and E. A. Stadelbacher.1991. Sex pheromoneproduction and calling in Helicoverpa zea (kpidoptera: Noctuidae): effect of tenrperatureand light. Environ.Entomol. 20: 145l-1456. Royer,L., and J. N. McNeil. 1991.Changes in calling behaviorand matingsuccess in the Europeancomborer (Ostrinia nubilalis),causedbyrelative humidity. Entomol. Exp. Appl. 6l:131-138. SASInstitute. 1996. SAS/STATuser's guide, version 6.1l. SASInstitute, Cary, NC. Webster,R.P., and R.T. Card6.1982. Influence of relativehumidity on callingbehavior of the femaleEuropean corn borer (Ostrinia nubilalis).Entomol. Exp. Appl. 32: l8l-185. Yel4 J. L., andM. Holyoak. 1997. Effectof moonlightand meterological factors on light and bait trapcatches of noctuidmoths (Lepidoptera: Noctuidae). Environ. Entomol. 26:1283- t290.

IN vol,.28 NO,2 SOUTHWESTERNENTOMOLOGIST JUN.2003

EFFECTS OF APHIDS, BARLEY YELLOW DWARF, AND GRASSY WEEDS ON GRAZED WINTERWHEAT

EmadA. Ismail,Kristopher L. Giles,Lisa coburn, Tom A. Royer,Robert M. Hunser. Jeanmarieverchot, Geraldw. Hornl, EugeneG. Krenzel, ThomasF. peeoel."' Mark E. Payon3,Gerald J. Michelsa,folhm S. Biblea,andDeb;;.b;idJ

Departnent of Entomologyand plant pathology, OklahomaState University, Stillwater, OK 74W S.

ABSTRACT

_ The effects of grazing winter wheat, Triticum aestiwtm L., on abundanceof aphids, barley yellow dwarf virus (BYDV), and grassyweeds (primarily chea! Bromus secalinus-!.), and on grain yield a1d yield componentswere determined during crop years 1999-2000and 2000-2001. cheat and aphidswere suppressedor enhaniedto lgguire a rangeof aphid and grassyweed infestationsin each-field. During the 1999- 2000 season,gopo'g reducedaphid abundance(as much as g7o/o)anrl ByDi levels (as much as 70o/o),but often promoted greater abundanceof cheai. stepwise regression analy.s99of 1999-2000growing seasondata indicated that grazingtrad iittle to no effect on yields or yield componentswhen aphidswere paf,t of the wheat system. In conhas! grailngwas significantly correlatedwith reduced yields when aphids were nor present during the 2000-2001season. Thesefindings may suggestthat the negativeeffects of grazing are not detectablewhen aphid abundanceis reduced as cattlJfeed on winter wheat.

INTRODUCTION

- Approdnntely 50Yoof the wheat planted in oklahoma is utilized as forage and p._g.uin production (dual-purpose). Greenbug,schizaphis graminum (Rondani), and birdcherry-oataphrd, Rhopalosiphum padi (L.), aretrvo oithe iost importantinsecipests of winter wheatin the SouthernGreat Plains. Theseaphids often reduceyields of small qr-a_iry,garticularly when infestationsoccur during earty plant growth stages(Burton et al. 1985,Kieckhefer and Kantack1988, Kieckhefer et al. t-99s,Riedell et;1. 1999,Kindler etal.2002). Additionally,both greenbugand birdcherry-oataphid hansmit barley yellow "w,hich d'warf virus @YDV) can further reduceyields (B".ton et al. 19g5, Massey 1993, Hunger a al. 1996,Hoffinann an! Kolb 1998). Riedeil et al. (1999) reported 216/ogran losscaused by birdcherry-oataphid, bfi 58o/oreduction ty tiroclerry-oa:taphid + BYDV.

ofAlimal science,oklahoma 'Deparfinent"Department stateuniversify, stillwater, oK7407g. of Plant and Soil Sciences,oklahoma state University, Stillwater, oK 74078. of Statistics,oklahoma lDepartnent-r,epartrnent stateunivercity, Stillwater,Stillwater, oK74o7g. of Entomology, Texas Agricultural Experiment Station, Bushland TX 79012.

t21 Dual-purposewinter wheat is often planted in early September,and grazing is terminatedin late winter (Winter and Musick 1991,Redmon et al. 1996). Planting wheat early often leadsto increasedgreenbug and birdcherry-oataphid intensifies (numberper tiller), and subsequentinfection by BYDV (Dedryrverand Tanguy 1984). Insecticide applicationsthat effectively reduceaphid intensity may have little effect on BYDV levels because most viral fransmission is thought to occur before aphid intensities reach economic thresholds (Hunger et al. 2001). Seed teatnents such as Gauchoil (imidacloprid) can inhibit aphid abundanceand subsequentBYDV levels, but are relativelyexpensive (Gourmet et al. 1996,Hunger et al. 1996). Plantingwheat early may promote infestationsby grassyweeds. Massee(1976) reportedthat downy brome is often more abundantin early planted wheat fields becausefinal tillage to preparethe seedbed, which dramatically reduces establishment of weeds, occurs before downy brome seedlingsemerge. Grazing winter wheat benefits cattle production in dual-purposesystems but seemsto increasethe abundanceof grassy weeds and reduce wheat felds. Dockage, which is primarily influenced by amountsof cheat,Bromus secalinusL., was 9o/ogeatsr in grazed versus nongrazedwheat (Koscelny and Peeper 1990 a, b). In conEastto the negative effects of grazhg winter wheat, Amold (1981) demonshatedthat grcnng reducesabundance of aphids. Grazingcould be potentially utilized to manageaphids and BYDV, which would reduce insecticide use and lower inputs. However, for dual- purposewheat, the effect ofreducing aphid abundance(via grazing) on grain yield has not been quantified. The tade-off between aphid and BYDV suppressionand the negativeimpact ofincreased cheatin grazedwheat are unknown. The objective of this researchwas to investigatethe potential of cattle grazingas a biologically based aphid/diseasemanagement approach over a broad range of dual- purposewinter wheat systems. The effects of grazing winter wheat on aphid intensity, BYDV levels, infestationsby grassyweeds, and the subsequentimpact of these factors on grain yield and leld componentswere evaluatedin severaldual-purpose wheat fields (with variable grazing intensities)in Oklahomaand Texasfrom 1999-2001. To evaluate the applicability of this aphid suppressionapproach and avoid purely experimental effects, care was taken to select a diverse set ofdual-purpose systemsrepresentative of the region. One important question relative to this researchwas whether aphid suppression, attributable to cattle grazingin wheat, resultedin yield protection. Yield protection via aphid suppressionmay be evident when aphids are reducedin grazedwheat, but yields are similar between grazed versus nongrazedwheat; i.e., yield reductions causedby aphids in nongrazedwheat are equivalentto yields n gra?,edwheat that are a firnction of the positive effect of removing aphids and the negative effect of gtazing. This idea would be further supportedif aphids were absentand yields were less in grazed versus nongrazedwheat.

MATERIALS AND METTIODS

llheat Field Locations. Experiments were conducted during the 1999-2000 growing seasonat three locations(near Perkins and Stillwater, OK and Sunray,TX), and at four locations during the 2000-2001growing season(Perkins and Stillwater, OI! and Bushland and Hartley, TX). Wheat fields near Perkins, Stillwater and Bushlandwere at Oklahoma State University and Texas Agricultural Experiment Station researchfarms, while the Sunray and Hartley fields were on private farms (Table l). Soil was analyzed for macro elements(N-P-K) at eachlocation to ensurethat adequateamounts were added before seedingfor a 3,360 kglha targetedyield goal in a dual-purposesystem' tn Adapted hard red winter wheat cultivars were drilled at 100 to 135 kg/ha into 20- cm rows (Table 1). Planting dates at each location were optimal for dual-purpose production (Krenzer 2000). With the exceptionof Perkinsand Stillwater during the first growing season,all fields were irrigated after planting to ensureseedling emergence. Expeimental Approach. T\e goal at each location was to establishexperimental units with quantitative measuresof gfazng (average leaf area), cumulative aphid intensities over time (aphid-days),BYDV index (incidence x severity), and infestations by grassy weeds (percentageweeds), and subsequentlyexamine the influence of these measureson wheat yields and selectedyield components. Limited land availability and number of cattle at each location, along with logistic difficulties associatedwith excluding cattle from a large number of independentplots, dictated that we arrange experimentalunits within grazedand nongrazed whole-plots (split-plot design). At each location,eight23 x 10.2-mwhole-plots were established within eachfield. Single-strand electricwire fencesexcluded cattle from the four randomlyselected nongrazed whole- plots. Four l0 x 3.6-msub-plots were established within eachwhole-plot. Eachsub-plot had a l-m border separatingit from other sub-plotsand the electric fence. Cattle often herdand excessivelyhample wheat near electric fences, To preventexcessive damage to grazed wheat, nongrazedand grazedwhole-plots were l0 m apart. For each location, sub-plotswere defined as experimentalunits for stepwiseregression analyses (see below) examining the influence of measuredvariables on wheat yields and selected yield components(M. E. Payton,personnel communication). This wasjustified basedon (1) the observation that cattle did not uniformly graze sub-plots within designatedgrazed whole-plots, (2) the observedrange of quantitative measurements(non-categorical) for all variablesamong sub-plots,and (3) measuredvariables were assessedidentically in all sub-plotstbroughout the study. Manipulation of sub-plotswithin each nongrazedor grazed whole-plot allowed establishmentof a quantitatively measurablerange of aphid-days, BYDV index, and percentageinfestation by grassy weeds. Within each whole-plot, cheat seedswere broadcastat 30 kg/ha in two randomly assignedsub-plots (Koscelny and Peeper1990b) after wheat emergencein 1999and beforeplanting in 2000. Natural infestationsby cheat were suppressedwith Maverickn herbicide(MON 37500) in sub-plotsnot receiving cheatseed. At Stillwater,hand hoeing among rows was usedto removeall othergrasses not confrolledby Maverick (mostly rye grass,Lolium species). During the 1999-2000 season,natural abundanceof aphidswas partially suppressedwith insecticidesin two randomlyassigned sub-plots per whole-plot(one with cheatseeding and one without). Malathionwas appliedbefore grazinginitiation (october-November)and Lorsban4E@ (chlorpyrifos, DowElanco) after grazingwas terminated(March); both were applied at recommendedrates. During the 2000-2001season, aphids were nearly absent;thus no insecticideswere used. Artificial establishmentof aphids(greenbug and birdcherry-oat aphid) was attemptedin desigrratedsub-plots, but was not successful. Grazing was initiated in November or Decemberand was terminatedin late Februaryor early March (Table l) before the advancementof the spike above the soil surface (at first hollow stem),where it would be removedby grazing@edmon et al. 1996). Measurements. Throughout the growing season,the following measurements were made in eacb sub-plot every 4-14 daysas weatherpermitted: tillers in two random lO-cm rows, aphid intensity (numberper tiller) including greenbug,birdcherry-oat aphid, English grain aphid, sitobion avenaeF., and com leaf aphid, Rhopalosiphumnaidis Fitch, on l0 randomly selectedtillers carefully clipped at ground level, and leaf areaof the same l0 tillers (measuredby a belt-fed leaf areameter). cumulative aphid-days(an aphid-day was defined as one aphid feeding on one tiller for 24 hours), were estimated under field conditions(Gerloff and hnan 1971,Pike and schaftrer 1985). Before r23 :sE2 in >l o CaI:! E d R 4 ."6 E: FRng 9:5

.o c: EF I Eg u E'3s i 3 F i a c.l d I iD rn C\l C) a rS c.l t I (t) * * r fEa * i N

cl tra a ,s i .. & s a h .HE N 0) Z S \o o\I F R n fi€. i\ h d' o\ o\ oo c ct l>l RatX6 td ltto q IE e,ZoZZrg E:() Frr6N6r-q lo F e.l C.l (\l H I it) o I x Pl - rI] Xl a) o0 e{t 6 *io At B rir rii J: A F N 6 g A !s el= = = -lu) NO\V I I I I o lo o if. 6- 0 tg gh F s E -aE l(l) o\ ii:: K = i- a.l tr

() d

€, !) €o0 Aoe t t .at F€ EE oo q) o g () B -j Eo o €) € E ,EE () oo -i g 6 9p F ? z .8".93* ru (t g F EE E t s s5

I'A harvest, percentageof infestation by grassy weeds (% weeds relative to wheat) was assessedin each sub-plot during tiller sampling (per l0-om row). Incidence of ByDV diseasewas determined using visual ratings as a percentageof plants that showed symptoms,whereas severity of symptomswas evaluatedon a scale of 0 to 9 (Qualset 1984),where 0 = diseasefree, and 9 = markeddwarfing, complete yellowing, few or no spikes with considerable sterility, forced maturity or dying of plants. An overall assessmentof the level of BYDV diseasewas estimatedby multiplying incidence by severity (all estimates were great€r than 0) and calculating a BYDV diseaseindex. During the 1999-2000season, conelations between calculated diseaseindex and both incidenceand severitywere significant(r > 0.8, P < 0.001). Thus, this diseaseindex seemsto standardizethe variability betweenincidence and severity measurementsand might be the most appropriateway to representthe overall effect of BYDV on wheat. During the 1999-2000se€tson, ELISA was used to confirm the presenceof the virus and identifr BYDV shains (PAV, RPV and RMV) in three symptomatic tillers from each sub-plotat eachlocation (Rochow 1979). During 2000-2001,BYDV wasnot detected. At maturity,height of threeor four randomlyselected wheat tillers was measured in each sub-plot. Before harvest,20 spikeswere randomly collected from each sub-plot to determineweight of kernelsper spikeand 100kemel weight. A small combinethat retainedboth cheat and wheat seedswas usedto harvestplants in a 13.4-m2area in each sub-plot. Harvestedkemels were cleanedwith a small seedcleaner to remove.cheat seedsso wheat yields and dockagecould be determined. Yields of wheat were adjusted to 13,5o/omoisture. Statistical Analyses. Data from each location for each year were analyzed separatelybecause (l) stockingrate variedamong locations, (2) wheatcultivars and the predominantaphid and weed speciesvaried among locations, and (3) preliminary regressionanalysis revealed significant effects of locationon yield (p < 0,001). Sub- plots at each location provided a broad range of quantitative measurementsof grazing (averageleafarea per tiller), aphid abundance(cumulative aphid-days), BYDV (disease index) and grassy weed infestation fuercentageweeds). Average leaf area,cumulative aphid-days, disease index, and percentageweeds from each sub-plot were used as independentvariables in a backwardelimination stepwiseregression model (pRoc REG, SAS Institute 1999). Yield (k/ha), plant height, tiller number,weight of kemelsper spike and lO0-kernelweight were includedas separatedependent variables. Because combinationsof measures(variables) may best describe dependentvariables, the stepwiseregression model included simple measures and all possiblecombinations. The significancelevel chosenfor stepwiseregression analyses was P: 0.05.

RESULTSAND DISCUSSION

Trendsfor Independent Yariables in RegressionModels. As expected, and similarto observationsby Koscelnyand Peeper (1990b), average leafarea was reduced at each location (as much as 460/o)in grazedsub-plots. During the 1999-2000season, birdcherry-oataphid and greenbugwere the predominantaphid species in oklahomaand Texas,respectively. Aphid intensities(number per tiller) at the Oklahomalocations were relatively low (<15), whereashigh levels(<45) were commonin wheatfields in Texas. Following introduction of cattle to experimentallocations, aphid-dayswere noticeably reduced(as much as 87%) in grazedcompared to nongrazedsub-plots (within sub-plots with no insecticidesand where cheatwas suppressed)at all locations. Similar results were observedby Amold (1981), who recommendedgrazing as a means of greenbug suppressionin wheat. During ttre 2000-2001season, aphids were nearly absentin the fai and spring and thus, it could not be determinedwhether their abundancewas affectedbv

125 grazrng, This low abundanceof aphidsmay be attributedto the late emergenceof wheat causedby severefall drought throughoutthe SouthemGreat Plains. Evaluationsby use ofELISA (1999-2000season) detected the presenceofthree BYDV stains: PAV, RPV and RMV, with PAV the most commonly observed. Rochow (1979) observeda similar prevalenceof PAV in field-+ollectedsamples from multiple statesduring a20-year period. Shain PAV was detectedin all sub-plotsat Perkins and many sub-plots at Stillwater and Sunray. The RPV strain was found only in a few sub- plots at Perkins,while the RMV shain was sparselydetected at all locations. Calculated BYDV index (incidence x severity) was markedly reduced(as much as 70%) in grazed sub-plots (within sub-plots with no insecticidesand where cheat was suppressed)at Stillwaterand Sunrayduring the 1999-2000season. At Perkins,calculated BYDV index measureswere similar amongall sub-plots.During the 2000-2001season, no symptoms of BYDV were observed. With the exception of Stillwater during both growing seasons,grazed sub-plots (with cheatseeding) at eachlocation had more grassyweeds (mainly cheat). At Perkins, grazedsub-plots (with cheat seeding)had as much as 15.4%omore weedsthan nongrazed sub-plots during both growing seruions;very small increasesin percentagesof weeds were observedat all other locations. At Stillwater, rye grasswas prevalentin nongrazed and grazedsub-plots during both growing seasons,but the prevalencewas decreasedby grazing. Using relativelyhigh stockingrates, Koscelny and Peeper(1990 a, b) reported an increasein cheat biomass causedby grazing. Grazing tends to remove the wheat canopy that shades slower developing cheat seedlings, thus allowing greater light penetrationand shifting the competitive advantageto cheat. More weedsin grazedsub- plots at Perkins may have been the result of the high gazing intensity (8.6 heads/ha) (Table l). As observedduring this study, the high stocking rate at Perkinswas sufficient to remove the wheat canopy. All other locations were moderately grazedwhich may have resultedin less light penetrationamong rows and consequentlyless growth of weeds. Step-llise Regression: Yield Components. At all locations, grazing (decreased average leaf area) was the most significant and often the only factor influencing (negatively)tiller numbers(F > 15.6,P < 0.001,Partial R' > 0.26). Reductionsin tiller numbers causedby gazing have been observedby several researchers(Shanow and Motazedian1987, Winter andThompson 1987 and 1990,Christiansen et al. 1989). Similar to previous findings (Pumphrey 1970, Koscelny and Peeper 1990b, Winter et al. 1990,), grazing (dpcreasingleaf area) resultedin shorter plants at maturity (F > 11.2,P < 0.001,Partial R' > 0.24)for all locationsexcept Sunray (1999-2000) and Perkins(2000-2001). At theselocations, grazing combined with aphid-daysand weeds negativelyinfluenced plant height at maturity(F > 7.5,P < 0.010,Partial R' > 0.26). Koscelnyand Peeper(1990b) found that wheat kernelsweighed less in grazed wheat at only one ofthree locations. In our study,the effectsof grazingon weight of kernelsper spikeand l00-kernelweight were variable. During the 1999-2000season at Stillwater and Sunray,weight of kemelsper spikewas not significantly influencedby any combinationof variables(P > 0.05). At Perkins (1999-2000),gru.ing, aphid-days, percentageof weeds,and BYDV index combinedhad a small but significantnegative influenceon weight of kemelsper spike(F : 6.4,P : 0.017,Partial R' : 0.18). During the 2000-2001season at Stillwater,Bushland and Hartley,weight of kemelsper spike w^asnegatively influenced by grazingor percentageof weeds(F > 8.9,P < 0.006,Partial R'> 0.22). At Perkinsduring the 2000-2001season, weight of kemelsper spikewas not significantly influenced by any combinationof variables(P > 0.05). During the 1999- 2000 season,100-kemel weight was not significantly influenced by any combinationof variables(P > 0.05). However,during the 2000-2001season, 100-kemel weight was

126 influenced grazingand/or legglivgtv by percentageof weedsat all locations(F> 4.7, p < 0.038,Partial^ff > 0.13). , step'wise Regression: Yields. Dnnng the 1999-2000season at perkins, aphid- days and percentageofweeds combinedwere the only variablesconelated with reduced yield (Table 2). At Stillwater and sunray, granng (leaf area), aphid-days,perc€ntage

TABLE 2. statisticaloutpy r91ltepwile RegressionAnalyses @ackward Elimination) for Simple and combined variables Relatedto yield (kgl ita) ai Each Location Durin! tr@

Parameter Partial Location Variable estimate F f 1999-2000 Perkins, Intercept 2209.7866 482.s9 <0.001 OK AD-PW -2.4323 20.36 <0,001 0.38 LA-AD.PW-BYDV -0.0001 9.99 0.004 0.19 AD-PW-BYDV 0.0041 t0.44 0.003 0.19 LA.AD.PW 0.0s23 20.09 <0.001 0.37 LA.PW 0.9357 8.22 0.008 0.15 Stillwater, Intercept 3t82.3667 174.55 <0.001 OK LA-PW -0.5610 2r.83 <0.001 0.40 AD-BYDV -0.2082 9.1I 0.005 0.17 AD-PW-BYDV 0.0044 7.20 0.012 0.13 Sunray, Intercept 4552.9479 36.67 <0,001 TX LA-PW-BYDV -0.0164 10.40 0.003 0.26 2000-2001 Perkins, Intercept 48t2.9336 1210.10 <0.001 OK PW 43.3103 82.12 <0.001 0.30 LA-PW 0.5691 10.00 0.004 0.04 Stillwater, Intercept 6785.6026 t224.60 <0.001 OK PW -70.4425 46.67 <0.001 0.28 LA.PW 0.9866 9.52 0.004 0.06

Bushland, Intercept 3720.1s74 101.38 <0.001 TX LA 52.334s 8.92 0.006 0.23

Hartley, Intercept 1317.1948 0.37 0.550 TX LA 24t.90/'2 8.54 0.007 0.22 LA: Averageleaf areaper tiller ( AD: Aphid days. PW = Percentageweeds/ sub-plot. = BYDV Barleyyellow dwarf virus diseaseindex (incidence x severity).

IN weedsand BYDV diseaseonly slightly affectedyield. Many studieshave demonsbated that aphid-dayscan be a useful predictor of grain leld loss at harvest(Kieckfieferet al. 1995, Kindler et al. 2002). In our study, however, aphid-days representeda minor component of the yield model at Perkins during 1999-2000. Interestingly, based on parameterestimates (Table 2), measuredvariables had little meaningfuleffect on yield at any ofthe locationsduring 1999-2000. During the 2000-2001season, aphid-days and BYDV diseasewere not detectable; thus, fewer variableswere includedin the stepwiseregression model. For both Perkins and Stillwater, the only variable greatly affecting (negatively) yield was percentageof weeds(Table 2). Basedon previouswork in winter wheat(Koscelny and Peeper 1990a), when other factorssuch as aphidswere not present,negative correlations between yield andabundance ofgrassy weeds are to be expected. Redmon et al. (1996) demonstratedthat increasing grazlu;Lgintensity reduced wheat yields. In our study, the only shong negative correlations between yield and grazing (averageleaf area) were detectedat Bushland and Hartley (Table 2). At both Bushland and Hartley, cheat was not well establishedand thereforedid not cause noticeably reducedyield. Basedon theseresults, it seemsthat in such wheat systemsas at Bushlandand Hartley, when weedsand aphidsare absentor scarce,negative effects of grazingare clearly demonstrated. Probably the most important question relative to this study was whether suppressionof aphids in grazedwheat resultedin yield protection. Yield protection via aphid suppressionmay be evident (l) when aphidsare reducedin grazedwheat, but yields are similar amonggrazed versus nongrazed sub-plots, and (2) when aphidsare not presentand yields are less in grazedversus nongrazed sub-plots. Both of theseideas are supportedby trends for yields observedduring this study. At eachlocation, aphidswere clearly reducedin grazedsub-plots during the 1999-2000growing season;during 2000- 2001,aphids were absent. Based on stepwiseregression analyses, gtazing (leafarea) had little to no effect on yields during the 1999-2000growing season when aphidswere part of the wheat system. In contrast,during the 2000-2001season, grazingwas significantly correlatedwith reducedyields when aphidswere not present' These findings may suggestthat the negative effects of grazingare not detectablewhen aphid abundanceis reducedas cattle feed on winter wheat, When greenbugsand birdcherry-oat aphids are cornmonin the late fall or early spring, grazing will reducetheir abundance.Based on our observations,when cattle are removed in the spring, aphid abundanceis unlikely to reach a damagingthreshold as wheat plants grow to maturity. Additionally, reducing aphid abundanceseems to decreasethe potential for developmentof BYDV. Therefore, grazinghasthe potential to be a biologically basedinsecVdisease management tactic within the integratedcrop and livestock wheat systemsof Oklahomaand Texas. However, ovet-grazingseems to promote competition from grassy weeds. Further study is neededto determineif moderate grazingin wheat systemsthroughout the SouthwestemU.S. holds promise for reducingpesticide inputs comrnonly used in wheatfields'

ACKNOWLEDGEMENT

We thank JasonKelley, Dennis Kastl, Adam M'clain, and SarahWedman for technicalassistance. We also thank RobertBarker and NathanWalker for their critical review of this manuscript.This work wasapproved for publicationby the Directorof the Oklahoma Agricultural Experiment Station and supported in part under projects OKI-02415 andOKL023 34.

128 LITERATURE CITED

Arnold, D. 1981. Effect of cold temperatureand grazingon greenbugpopulations in wheat in Noble countn oklahoma, l97s-76, J. Kansas-Entomo'i.-soc.s+i szr- \TT Burton,R. L., D. D, simon, K. J. stark, and R. D. Morrison. 19g5. seasonaldamage by greenbugs (Homoptera: Aphididae) to a resistant and a susceptiblevariety of wheat.J. Econ.Entomol. 78: 395401. phillips. christiansen,s., T. Svejcar, andA. 19g9.Spring and falr cattlegrazing effects on componentsand total grainyield ofwinterwheat. Agron.r. st: l4s-tsb. Dedryver, c' A., and S. Tanguy. 1984,Biorogy of cerealaphids in ttrewest of France.v. Effect of sowing date of winter wheat on the infestation of fields by Rho-palosiphumpadi (L.), sitobion avenae(F.) and Metoporophiumdirhodum (wlk.) and on the developmentof aphid populationsduring the rolowing spring. Agronomie4: 7ll-719. Gerloff' E' physiological D.' and E. E. ortman. 1971. changesin barley inducedby greenbugfeeding stress. Crop Sci. I I : 174_17 6. Gourmet, pedersen. c., F. L' Kolb, c. A. Smyth,and w. L. 1996.Use of imidaclopridas seed-teatment insecticide to control barley yellow dwarf virus (ByDV) in oat andwheat. Plant Dis. 80: 136-141. Hoffrnann, T. K, and F. L. Kolb. 1998.Effects of barleyyellow dwarf virus on yield and yield componentsof drilled winterwheat. plant lis. g2:620_624. Hunger, R. p. M., J. L. Sherwood,G. Krenzer, Mulder, and M. E. payton. 1996. Evaluationof Gaucho480F seedtreatment to control aphidsand barleyyellow (BYD) dwarf in hard red winter wheat.Fungicide and NenrLaticideTests 52:326. Hunger, R. M., B. Olsen,_W.C. Siegerist,E. G-,Krenzer, and M. E. payton. 2001. Evaluation of Gaucho 480F seedteatrnent to contror aphids and barley yellow ,dwarf@YD) in hard red winter wheat (HRww). Fungicideand Nematicidi Tests - _ Vol. 56 (http//www.scisoc. org/online.F&Ntests/). Kieckhefer,R. w, and yield B. H. Kanta;k. r9gg. lossesin winter grainscaused by cerealaphids (Homoptera: Aphididae) in South Dakota.J. Econ. Entomol. g1: 3r7-32t. Kieckfiefer, R. w., J. L. Gellner, and w, E. Riedell. 1995.Evaluation of the aphid-day as predictor {ggdard a ofyield loss causedby cerealaphids. Agron. r. az, zss- 788. Kindler, s. D', N. c. Elriott,f( L. Giles,T. A. Royer,R. Fuentes-Grandaos,and F. Tao 2_002' Eff:ct of grgenbug(Homoptera: Aphididae) on yield loss oi-winter wrreat. J. Econ.Entomol.95: 89-95. Koscelnl J. A., and r. peeper. F. 1990 a. Herbicide-grazinginteractions in choat (Bromussecalinus)-infested winter wheat(Triticum-aestiv"um). WeeJ-s"ience f S, 532-535. Koscelny'J. A., peeper. and T. F. 1990b. Effect of grazingcheat (Bromus secalinus)- infested wheat(Triticumaestiwn) on both sp..ier. d"ed recir. q: sol-soa. Krenzer,E. 2000. wheat as fo_rage,pp. 27-30.rn i. A. Royerand E. G. Krenzer wheat management 1eds.) in oklahoma. oklahoma coopeiative Extension service. E- 831. Massee,T. w. 1926. Downy bromecontrol in drylandwinter wheatwith stubblennrlch fallow and seedingmanagement. Agron. J. 6g: g52-gSS. Massey, B. 1993. Insectson smalr grains and their control. oklahoma cooperative ExtensionService No. 7176.

t29 Pike, K. S., and R. L. Schafftrer, 1985.Development of autumnpopulations of cereal aphtds,Rhopalosiphum padi andSchizaphis graminum and their effects on winter wheatin WashingtonState. J. Econ.Entomol. 78: 676-680. Pumpbrey, F. V. 1970. Semidwarf winter wheat responseto early spring clipping and grazing.Agron. J. 62:641-643. Qualset,C. 1984.Evaluation and breedingmethods for BYDV resistancepp. 72-82.ln P, Burnett(ed.) BYD, CIMMYT, Mexico. Redmon,L. A., E. G. Krenzer,D. J. Bernardo,and G. W. Horn. 1996.Effect of wheat morphologicalstage at gazingtermination on economicretum. Agron. J. 88: 94- 97. Riedell, W. E., R. W. Kieckhefer,S. D. Haley, M. C. Langham, P. D. Evenson.1999. Winter wheat responsesto birdcherry-oat aphids and barley yellow dwarf virus infection.Crop Sci.30: 158-163. Rochow, W. F. 1979.Variants of barley yellow dwarf virus in nature.Detection and fluctuation during twenty years.Phytopathology 69: 655-660. SAS Institute.1999. SAS/STAT guide for personalcomputers. 8.1 ed. SAS Inst.,Cary, NC. Sharrow,S. H., and L Motazedian.1987. Spring grazirLg effects on componentsof winter wheatyield. Agron. J.79:.502-504. Winter, S. R., and J. T. Musick. 1991.Grazed wheat grain yield relationships.Agron. J. 83:130-135. Winter, S. R., and E. K. Thompson.1987. Grazingduration effects on wheatgrowth and grainyield.Agron. J.79: 110-114. Winter, S. R., and E. K. Thompson. 1990' Grazing winter wheat: I. Responseof semidwarfcultivars to grain and grazedproduction systems. Agron. J. 82:33-37. Winter, S. R., E. K. Thompson,and J. T. Musick. 1990.Grazing winter wheat:Height effecton responseto productionsystem. Agron. J. 82:3741.

130 vol.28 NO.2 SOUTHWESTERNENTOMOLOGIST JuN.2003

THE NEGATIVE BINOMIAL DISTRIBUTIONAS A MODEL FORDESCRIBING COUNTSOF GREENBUGS, SCHIZAPHIS GMMINUM,I ON WHEAT2

N. C. Elliott, K. L. Giles3,T. A. Royer3,S. D. Kindler,D. B. Jones3,and F. L. Tao3

USDA, ARS, PlantScience Research Laboratory, N. WesternSt., Stillwater,OK 74075

ABSTRACT

One-hun&ed-thirty-two samples in which one or more tiller was infested with greenbugs,Schizaphis graminum (Rondani), were taken during the fall growing senson, and 232 sampleswith one or more greenbugswere taken during spring. The frequency distribution of greenbugcounts differed significantly from that expectedfor a negative binomial distributionfor six of 132 samplesfrom fall (4.5%) and for ix of 222 samples from spring(2.7%). Estimatesof ft werehighly variableand changedsystematically as 7 increased. The parametersof a quadraticregression model fitted to sampleswith t > 0.25 greenbugsper titler did not differ significantly betweensamples from fall and spring. The expectedvalue of ft from the quadraticregression model varied from approximately0.3 to 0.7 as F varied from 2 to 12 greenbugsper tiller. Variationin estimatesof t aboutthe regressioncurve providedan indicationof the extentto which ft varied independentof greenbugpopulation intensity. The meansqrure error of the regressionwas equalto 0.043 and was relatively large comparedto t. This indicatesthat a sequentialsampling scheme basedon the negativebinomial distribution should be evaluatedby simulationand/or direct validation to veriff the utility of a samplingscheme before operationaluse in a pest managementprogram.

INTRODUCTION

Each year, more than two million ha of winter wheat, Triticurn aestivum L., are planted in Oklahoma. The greenbug,Schizaphis graminum (Rondani),commonly infests winter wheat in Oklahoma and causessevere damage, with estimatesof economiclosses exceeding$100 million in someyears (Webster 1995). A sequentialsampling scheme basedon binomial count (presence-absence)sampling of greenbugson individual wheat stems(tillers) was recently developedfor the greenbugin winter wheat (Giles et al. 2000). This sampling schemewas derivedfrom a commonlyused empirical equation relating the

'Homoptera:Aphididae zMentionof tradenames or commercialproducts in this articleis solelyfor thepurpose of providing specific information and doesnot imply recommendationor endorsementby the U.S. Deparfnentof Agriculture. rDepartrnentof Entomology and Plant Pathology,Oklahoma State University, Stillwater, oK74078

131 meannumber of insectsper 'nir sample to the proportionof infestedsample units (Kono and Sugino1958, Gerard andchiang pzo;. wirle trris;il#;;ffi';- be useidto developsequential sampling schemel tnoreuur"d on (1947i;;q;;;ial probability ratio test (SPRT) are sometimes, precise. fald;s more This is becau.eil i, ;;;;"..rsary with sPRT to incorporatevariance of the regression relationship ;iih; ;r*;", of sample units infestedwith the particular pest to ihe meannumber- of pestsper sampreunit (Nyrop and Binns 1991)' To deve^lopa iampling schemeusing the Spnf, it n,"rile known that the distributionof countsof insectsin sariptesfits a part'icularp.ouuuilitv Jirtribution. For insects such "s_the greenbug, which have aggregatedspatial distributions, the negative binomial distribution sometimesprovides ii "i""ptuui"-fir;' sample (Southwood1978). data The negativebinomiar distribution_is "il";bri;;i;y ilo p**"r"rr, the mean_andan exponentdenoted by,t (Southwood l97g). wh.;;il'si'rtr ro derivea sequentialsampling scheme based on the negativebinomial "rid sp;i}, ; is assumedto be constant. Failure of /r to remain .onrtiot complicatesa"r!rop*"rt "f a sampling because 9chem3 the probabilities of rejecting the hypother", thui-;;;;birg population intensity is less than or greater than an actioir threshold differ from the nominal values establishedin derelopingthe _scheme(Nyrop and Binns r99l). irr" .oiu.tn..s of the samplingplan lequgnli{ may dependon how mucht variesat densitiesi"* trt. economic thresholdfor the insect. The two objectivesof this studywere to o"te.mine *t "ther: 1) the negativebinomial distribution provideda good fit to data collecteJtoi g.""nuug, from winter wheat fields in oklahom4 and 2) f varied ov_erthe range of vJue! or tne sampre mean (7) t)?ically observedin greenbug-infestedfields, uno-r it diffe;d, to describe mathematicallythe form of the relationshipbetween k and x .

MATERIALS AND METHODS

From September . 1997rhrough May 200r, the numberof greenbugsper tiller was estimatedin fields of hard red winter wheat located through;ut ..nfra'urro western oklahoma"including the.panhandle- rields sampleafrom seiteJ"ilhr*gr, December were considered to have been sampledduring the fal growinj s"*orL *J iirose sampred from Januarythrough May were considered-tobe fieiis ,"-"pr"J-J*irg1f,ring. Tillers were sampledfrom a field by walking a u-shapedtransect through "rr-u"i a fie'id ano cutting at groundlevel an individual tiller approximatelyevery l0 m. The "r gr.enbugson eachof 80 to 200 tillers was countedin the field and recorded. The numberof tillers inspectedin_ a field dependedon greenbugpopulation intensity, *ith more tilers being inspectedwhen_greenbugs were scarcethan when they were-abundant. In total, 132 samplesin which oneor moretiller was infestedwith greenbugswere taken during the fall growingseason' and232samples-with one or moregreinbugs iere takenduring spring. /r of the negative binomial distributionwas estimatedby using ttre maximum likelihood methoddescribed by Bliss and Fisher(1953) for eachsannple. ir," e";ii"* of fit to the negativebinomial distributionwas determinedfor each.*rpt" uy ;irg; ;hi-rquaretest and Fisher @liss 1953). The probabilityof rejectingthe hypb*resisttraittre observeddata fit the negativebinomial = was set at o 0.05. Analyseswere accomplishedwith a computerprogram written in CompaqVisual Fortran 6.6@. relationshipbetween t and the _The mean number of greenbugsper tiller ( f ) in the sampleswas -investigatedby regressionof /c on f . A previoui siudy indicatedthat parametersof a statistical 1of] relating the population intensity of greenbugsto the proportionof tillers infested greenbugs, -with whictr implicitly a#naJon the-negative binomial distribution, differed for wheatfields sampleda*ing rai .ornp-.0 with fields

r32 sampled in spring (Giles et al. 2000). Therefore, we used heterogeneityof slopes regressionto test for a difference in the relationship of & to t betweenfall and spring. Regression analyses were accomplished using the REG procedure of the Statistical Analysis System(SAS Institute 1990). A quadraticregression model with a single linear and quadraticparameter for data from both seasonswas also fitted.

RESULTSAND DISCUSSION

Maximum likelihood estimatesof the parameterft variedfrom 0.002and 145.1for fields sampledin fall, andfrom 0.001to259.2 for fields sampledin spring. Estimatesof ft were highly variable when the mean number of greenbugsper tiller (7) was less than approximately0.25 for fields in the fall and spring (Fig. l). The value of t increased slightly as F increasedfor greenbugsin fields sampled in fall and spring and then decreasedas 7 increasedto more than 15 greenbugsper tiller.

fall

a oo a

*€ rt""'c' E -8 ED 38 E6 spnng 3 I (U4 I z I

o.9 t

MeanNumber of GreenbugsPer Tiller

FIG. 1. Natural log of maximum likelihood estimatesof the parameterrt versusthe mean numberof greenbugsper wheattiller for fields sampledin fall and spring.

r33 ^ Th. frequency distribution of counts of greenbugsper tiller differed significantly from that expected for a negativebinomial protiUitity distributionfor slx oit:n+ .56/0);f the fields sampled in fal and for six of 222 (2.7%) ;f the fields sampledin spring. These percentageswere only slightly less than the percentageof samplis expectedto differ significantly by chancegiven-that .' was set equal to 0.05 for each chi+iuare test. The general agreementofobserved to expectedfrequencies indicates that the nelative binomial is an acceptableprobability model for describingthe frequencydistributiJn of counts of greenbugson wheattillers for fields sampledin the fall or spring. Becauseofthe apparentnonJinearity in the relationshipolt lo i, aquadraticregression model was used to fit the relationship, sampleswith t< 0.25 were excludedfor t}le purpose of modeldevelopment because of the greatvariability of the estimatesof &. Linear and quadratic termswere significantfor fields sampledin att ano spring(Table l). All regressionparameters were significantly different from zero, but i'regression of heterogeneity of slopes indicated there were no significant difierences for any of the parametersbetween fall and spring. -Even thoughparameters ofthe quadraticregressions differedfor fall and springsampled fields, the differenceswere not significant.Tirerefore, fall and springdata were combined to constructa singleregression .id.l. p*u^eters for linear and quadraticterms were significant for the model using the combined data. The modelexplained 48% of thevariability in ft.

TABLE l. Parametersof QuadraticRegression Models Relatingthe Parameter/c of the NegativeBinomial Distributionto the MeanNumber of Greenbugsper Tiller (i) for Fall andSpring Sampled Winter Wheat Fields, and for All SampledFi;lds.

n lntercept Slope Slope2 Fall 50 0.12(0 0.060(0.014) -0.00r4(0.0007) 0.54 Spring 95 0.20(0.099) 0.088(0.014) -0.0015(0.0007) Both Seasons I 45 0.20(0.02s) 0.062(0.01l) -0.00r5(0.0006) 0.48 "Standard errorsofestimates are in parentheses.

Agreementof frequencydistributions of greenbugcounts on wheat tillers with expectations of the negative binomial distribution doei not assure that the negative probability 9ryfi4 distributionconectly describesthe observedprobabilities (Hurlbert 1990); However,from a practicalperspective, the small numberof significantdeviations indicatesthat at the very least,the negativebinomial distribution providis a goodempirical probability model to describethe distributionof counts. Undlr this circumstance,the negativebinomial is acceptablefor developingsequential sampling schemes by usingthe SPRT(Binns 1994). The systematicvariation of & with f observedwill complicatedevelopment of sequentialsampling schemes for greenbugs.The extent of this problem will dependon the amount of variation in /r over the range of mean population intensity for which control decisionsare typically made (Nyrop and Binns l99l). The valueof Lseemsto vary from about 0.3 to 0.7 for - varying from approximatelytwo to 12 greenbugsper tiller (Fig. 2), th9 rangeof population intensitiesfor which control decisionstypUty ure madein winter wheat in oklahoma (Royer et d. l99s). Thus, variation in the value of /r with respectto greenbugdensity is not exhemeover the rangeof 7 for which control decisionsare made. Variation aroundthe regressioncurve (Fig. 2) indicatesthe extent to which estimates oft can vary independentofgreenbug population intensity. The mean squareenor ofthe regressionis a good estimateof the potential maximumof thisvariabiiity(Nyrop and

tu 1.2

1.0 o p eo^l ao ct 0.8 ao a , o.-7a o. 0.6 @o * Ir

fall 0,2 % o'o o b-'. ospring 0.0 051015202530 Illean Numberof GreenbugsPer Tiller FIG.2 Linear regressionequation for & versus 7 for wheat fields sampledin fall and spring.Sampleswith t < 0.25were excluded from the regressionanalysis.

Binns l99l). The mean squareenor for the regession(M,SE : 0.043) was large as a proportionof /c,ranging from 0.14 for t : 0.3 to 0.06 for k : 0.7. An unknownfraction of the variation in t is causedby samplingerror in the estimate,and will haveno effect on the performanceof a sequentialsampling schemeapart from its influence on the precision of the estimateof /cused in developingthe plan. However,true variationin ft independentof x is probablyalso present,which will affect the performanceof the samplingscheme. The largevariation in /r indicatesthat a samplingscheme based on the negativebinomial distributionshould be evaluatedusing Monte carlo simulation(Nyrop and Binns 1991) and/or direct validation using independentsampling data to accuratelyassess the utility of the samplingscheme before using it to makegreenbug control decisions.

ACKNOWLEDCMENT

This researchwas partially supportedby a grant from the oklahoma center for Applied Scienceand rechnology. we thankB. Irvin, A. Mackay,R. Fuentes,T. Johnson, w. French,J. spurlin, andJ. Shelbumefor technicalhelp. we alsothank wade Frenchand Phil Mulderfor their criticalreview of this manuscript.

LITERAruRE CITED

Binns' M. R. 1994. Sequentialsampling for classifringpest status,pp. 137-174.InL. Pedigoand G. D. Buntin [eds.]Handbook of samplingmethods for arthropodsin agriculture.CRC, Boca Raton, FL. Bliss,C. I., andR. A. Fisher1953. Fittingthe binomialdistribution to biologicaldata and a noteon the efficientfitting of the negativebinomial. Biometrics9:176-200. Genard, D. J., and H. c. chiang. 1970. Density estimationof com roorwonn egg populationsbased upon frequency ofoccurrence. Ecology 51:237-245.

135 Giles, K. L., T. A. Royer,N. C. Elliott, and S. D. Kindler. 2000. Developmentand validation of a binomial sequentialsampling plan for the greenbug(Homoltera: Aphididae)infesting winter wheatin the southernplains. J. Econ.Entomol.g3:'1522- I 530. Hurlbert,S. H. 1990. Spatialdistribution of themontane unicom. Oikos 58: 257-262. Kono, T., and T. Sugino. 1958. On the estimationof the densityof rice steminfested by the rice stemborer. Japan.J. Appl. Entomol.Zool.2: 14-188. Nyrop, J. P., and M. Binns. 1991. Qtrantitativemethods for designingand analyzing samplingprograms for use in pest management,pp 67-132. /z D. pimentel[ed.], Handbookof pestmanagernent in agriculture:Volume II. CRC Press,Boca Raton, FL. Royer, T. A., K. L. Giles, and N. C. Elliott. 1998. Small grain aphidsin Oklahoma, OklahomaCooperative Extension Service, Oklahoma State University Extension Facts.Sheet no. FS-7183. SAS Institute. 1990. SAS/STATUser's Guide, Venion 6, FourthEdition. SAS Institute, Cary,NC. Southwood,T. R. E. 1978. Ecologicalmethods with particularreference to the studyof insectpopulations. Chapman and Hall, London. Wald,A. 1947. Sequentialanalysis. John Wiley andSons, New York. Webster,J. A. 1995. Economicimpact of the greenbugin the westernUnited States: 1992-1993. PublicationNo. 155. Great Plains Agricultural Council, Stillwater, Oklahoma.

t'x vol,.28 NO.2 SOUTHWESTERNENTOMOLOGIST JuN.2003

PROTECTIONOF THE SPORE-TOXINCOMPLEX OFB A CILLUSTHUNNGIENSIS SEROVAR/,SRIEIENSIS FROM ULTRAVIOLET IRRADIATION WITH ALUMINUM{MC ENCAPSULATIONAND PHOTOPROTECTORS

Ramirez-LepeM, O. Aguilar,M. Ramirez-Suero,and B. Escudero InstitutoTecnol6gico de Veracruz.Unidad de Investigaci6ny Desarrolloen Alimentos ApartadoPostal 1380. C.P. 91860 Veracruz, Ver. M6xico. E mail:[email protected]

ABSTRACT

A Bacillus thuringiensisserovar israelensis (B.r.i) spore-toxin complex was encapsulatedwithin aluminum carboxymethylcellulose(cMC) using green malachite, ponceaured, and congo red photoprotectiveagents against ultraviolet light. Two percent CMC was found to be a stableencapsulated spore toxin-complex. The LCsoand LCesvalues of the spore-toxincomplex were increased13.9 and 29.lyo, respectively,and slope decreased13.6% through the microencapsulationprocess. when malachitegreen, ponceau red, andcongo red were added,the LCsochanged from 0.6155mg/l ofspore-toxincomplex encapsulatedto 0.6127,0.6207and 0.6738 mgA, respectively. Twenty-four hour exposureof the free spore-toxinto ulbaviolet light reducedstability to zero as indicatedby testsagainst third-instarAedes aegtpti (L.) larvae.When the spore-toxincomplex was microencapsulated within aluminumcarboximethylcellulose alone and with ponceaured, malachitegreen, or congored as photoprotectoragents and then exposedto uv light for 24 hrs, percentage mortalityagainst third-instar le . aegyptilarvae was 23, 50,60 and 88%, respectively.The aluminum-CMCencapsulated form of the sporetoxin complexof B.t.i. with photoprotectors avoidedthe limitation of this agent in controlling mosquitolarvae causedby ultraviolet light.

INTRODUCTION

Bacillusthuringiensis serovar israelensis (B.t.i), a gram-positivebacterium, is highly toxic to dipteranlarvae such as mosquitoesand blackflies which are vectors of several tropicaldiseases as malariaand dengue(de Barjac 1978).The entomocidalactivity of B.t.i. is due to crystalline inclusionsproduced during sporulation.Inclusions are ingestedby susceptiblelarvae, solubilized by the high pH of the larval midgut, and the protoxin is activatedby proteases.The gut cells swell and lyse, and the insect stops feeding and eventuallydies (Ellar 1994).The use of B.l.i. is limited by the low efficacy of current preparationsunder field conditions(Tyanyun and Mulla 1999).Major reasonsfor the low efficacy of thesepreparations include particle sedimentation(Rushed and Mulla 1989), proteinadsorption onto silt particlesand organic matter (Ohana et al. I 987),consumption by other organismsto which the toxin is not lethal (Blausteinand Margalit l99l), dissolved tannins(Lord and Undeen 1990),and inactivationby sunlight (pusztaiet al. 1991). Photoinactivationseems to be one of the major environmentalfactors affecting the stability of B.t.i.deltaendotoxin(Leongetal. 1980). Inothersubspecies of B.thuringiensisactive againstlepidopterans, polymeric formulations and photoprotective agents havJbeen used to encapsulatethe spore-toxin complex to protect the O-endotoxinfrom environmentalfactors (e.g.,Tamez-Guerra et al. 1999,McGuirre et al. 1996,Behle et al. 1996,Ridgway et al. 137 1996, Castro-Francoet al. 1998). However, theseformulations cannot be used against dipteran larvae becausetheir behavior, and habitat needs are different from those of lepidopteranlarvae. In the subspeciesisraelensis, some studies have been reported on the protectionof the deltaendotoxin against environmental conditions (e.g., Kase and Branton 1986,Yu-Tien et al. 1993). In this paper,we presentthe resultsof a study on the photoprotection of mosquitocidal activity of a spore-toxin complex with carboximethylcelluloseand aluminum sulfate using malachite green, congo red, and ponceaured as photoprotectiveagents.

MATERIALS AND METHODS

B.l,i. was obtainedfrom the PasteurInstitute, Paris, France.An active culture was maintainedby steak-inoculating nutrient agar-slanttubes, incubated at 30'C for 2 days,and storedat 4'C. The culture was transferredto new agarslants at approximatelytwo-month intervals.Two milliliters of LB medium(1.0% tryptone, 0.5% yeast extract and 1.0%NaCl) were inoculatedwith 50pl of suspensionof grownagar cells (O.D.ezo=2.0 approximately). The culturewas incubatedovemight at 30"C and then transferredto 50ml of mediumin a 250-ml flask. The compositionof the medium(expressed as g/l in distillatedwater) was as follows:tryptone, 15; Na2HPO4,7.l; MgSOa7HzO,0.02; FeSOa7HzO,0.001; ZnSOaTH2O, 0.005;and CUSO+5HzO, 0.005 with pH=7.0* 0.2 (Pearsonand Ward 1988).The culture wasincubated in a shakerat 300 rpm and30oC and harvested at 48 hrs whenlysis reachedat least90%. The fermentedbroth was centrifugedat 5,000 rpm for l0 min, washedthree timeswith 0.1 M NaCl, and then with distilledwater. The spore-toxincomplex was dried at 60"C for 48 hrs. Solutionsof al.0Yo (w/v) spore-toxincomplex were madeup in sterile water containing0.5-3.0% (w/v) sodium CMC (Sigma)to which 2.0% (w/v) of the respectivephoto protectiveagent was added(i.e., malachitegreen, congo red, or ponceau red). The solutionswere magnetically stined ovemight at 4oCand then transferred into 0.05 M Alz (SOa)I (pH=3.4)as describedby Elcin et al. (1995)using a hypodermicneedle at a drop rate of 60 drops per minute. The mixture was stirred at 4'C until aluminum CMC microcapsuleswere formed. Microcapsuleswere filtered through a sterile filter (MF millipore 1.2 pm), washedseveral times with sterile0.05M Alz(SOa)r(pH=3.a), and then lyophilized. Proteincontent of the microencapsulatedsamples were assayedby the Lowry et al. (1951)method. Bioassayswere carriedout accordingto the test proceduresobtained from Mclaughlin et al. (1984). Lawae of Aedesaegtpti (L.) wereobtained from colonies maintainedat the insectaryof Ministry of Healthof Veracruz,Mexico. Fifty milligramsof the spore-toxincomplex or microencapsulatedpowder were placed in 5ml of deionized water.Stock solutions were preparedin test tubesby serialdilutions of the homogenatesin deionizedwater, and appropriate dilutions were tested against third-instar ledes aegtpti (L.) larvae.Five different concentrationsof eachpowder were preparedin duplicatefor each assay.In eachassay, 20 mosquitolarvae were addedto a final testvolume of l00ml. Each testwas repeatedthree times on differentdates. Duplicate cups with 20 mosquitolarvae in 100m1of deionized water without test material served as a control. Mortality was determinedat 24 hrs,based on the numberof deadlarvae. Mortality was subjectedto probit analysis(Finney 7917) on a computerprogram and the LCso,LCqo and curve slopewere calculatedfor the respectivespore-toxin complex and microencapsulatedpowder. Efficacy of formulations,LCso, LCgo,and slopesvalues were comparedby analysisof variance (p<0.05)using a Minitab@ 10.2statistic program. Stability of free and encapsulatedspore- toxin complexmixtures againstultraviolet light was testedusing an UV germicidallamp (257 nm). Sampleswere suspendedin 100m1of deionizedwater and exposedto UV irradiationat a distanceof 60cm. Duplicatecups wdre collected at 6, 72, 18 and24 hrs,and

138 Table I showsthe probit analysisof the concentration-mortalityresponse of third- instar Ae. aegtpti larvae to the encapsulatedproducts. These results indicated that LC5e valuesfor spore-toxincomplex were 15.9%olower than that for the encapsulatedproduci without the photoprotector,and 15.4, 16.9 and 26.9%lower for the encapsulatedpioducts with malachitegreen, ponceau red, and congored, respectively.However ic56 valueswere not significantlydifferent for the unencapsulatedtoxin and the encapsulatedtoxin with and without a photoprotector.The highestLCee value for the encapsulatidproducts was'for the congo red formulation (1.6218mg/l), and the lowest was for the formulationwithout a photoprotector(l.3872 mg/l). All encapsulatedproducts showed an increaseof 29 to 52yo in the LCqovalue with respectto the free spore-toxincomplex formulation. The LCesvalue was significantlydifferent for encapsulationwith ponceauand congo red with respectto the frce toxin. The slopes ranged from 13 to 2\yo less in value for the encapsulated formulationscompared to that for the unencapsulatedspore-toxin complex; here we also observedvalues significantly different for ponceauand congo red.

TABLE I . Toxicity (LC5eand LCee,mg/l) of MicroencapsulatedFormulations of the B.r. i. Spore-ToxinComplex to Third lnstu Aedesaegtpti LarvaeOver 24 Hours.

PRODUCT LC5s +/- s.d. LCee+/- s.d. SLOPE+/-s.d.

Spore-toxincomplex 0.53104+/-0.0266 1.0740a+/-0.0835 4.2728a+t-0.5840

Encapsulation without 0.6155b+l-0-0473 1.3872ab+/-0.13883.6916ab +/-0.2637 photoprotector

Encapsulation with 0.6t27ab+t-0.05971.4488ub +L0.37g6 3.627tac+/-0.6715 malachite green

ab+/-0.0330 Encapsulationwith 0.6207 t.$42b i-0.2648 3.0+t6bt+/-0.5049 ponceau red

ab+/-0.0788 b+i-o.1655 Encapsulation with 0.6738 1.6218 3.4002b"+/-0.4183 congo red

'Different letterson a given determinationindicate significant differences among formulations by one fitctoranalysLs of variance(p(0.05).

Fig. 3 shows the effects of ultraviolet irradiation on the stability of the encapsulated and normal spore-toxin product. These results indicate that UV inadiation affected the activity of the unencapsulatedproduct to a point where, after 24-hrs inadiation, less than 5% mortality was realized when the inadiated formulation was tested against third-instar Ae. aegtpri larvae. The aluminum-cMC encapsulated spore-toxin compiex without a photoprotectorwas also affectedby r.rltravioletirradiation, resr.rlting in less than 25Yo lawal mortality after inadiation for 24 hrs. Results of the larvicidal activity test involving the aluminum-CMC encapsulatedformulations containing the photoprotectors malachite green or ponceau red demonstratedthat major protection ofthe spore toxin against ultraviolet light w-asrealized, with these particular formulations still causing 55o/oor more larval mortality after 24 hrs of inadiation. In the case of the congo red encapsulatedformulation, almost thL same larval mortality rates (88%) resulted after irradiation as before (Fig 3).

t39 Table I shows the probit analysis of the concentration-mortality response of third- instar Ae. aegtpti larvae to the encapsulated products. These results indicated that LCso values for sp6re-toxin complex were 15.9% lower than that for the encapsulatedproduct without the photoprotector, and 15.4, 16.9 au:d26.9% lower for the encapsulatedproducts with malachite green, ponceau red, and congo red, respectively. However LC59 values were not significantly different for the unencapsulatedtoxin and the encapsulatedtoxin with and without a photoprotector. The highest LCe6value for the encapsulatedproducts was for the congo red formulation (1.6218 mg/l), and the lowest was for the formulation without a photoprotector (1.3872 mg/l). All encapsulatedproducts showed an increase of 29 to 52o/o in the LCqo value with respect to the free spore-toxin complex formulation. The LCqo value was significantly different for encapsulation with ponceau and congo red with respect to the free toxin. The slopes.ranged from 13 to 28Vo less in value for the encapsulated formulations compared to that for the unencapsulatedspore-toxin complex; here we also observed values significantly different for ponceau and congo red.

TABLE L Toxicity (LCso and LCqo, mg/l) of Microencapsulated Formulations of the B.t i. Spore-Toxin Complex to Third lnstu Aedes aegpti Larvae Over 24 Hours.

PRODUCT LC5e +/- s.d. LCqo+/- s.d. sLoPE+/-s.d.

Spore-toxincomplex 0.53104il-0.0266 1.07404+/-0.0835 4.27284+/-0.5840

Encapsulation without 0.6155b+^0-0473 1,3872ub+/-0.1388 3.6916ab+l-0.2637 photoprotector

ab+l-0.0597 Encapsulation with O.6t2i 1.4488ab+t-0.3796 3.62784c+l-0.6715 malachitegreen ab+/-0.0330 Encapsulation with 0.6207 t.$42b +t-0.2648 3.04160c!-0.5049 ponceau red

ab+/-O.0788 b+/-0.1655 Encapsulation with 0.6738 1.6218 3.4002bt+t-0.4183 congo red

Diff*nt l"nert on a given determinalionindicate significant differences among formulaiions by one l?lctoranalysis of variance(p<0.05).

Fig. 3 shows the effects of ultraviolet irradiation on the stability of the encapsulated and normal spore-toxin product. These results indicate that UV inadiation affected the activity of the unencapsulatedproduct to a point where, after 24-hrs inadiation, less than 5% monality was realized when the inadiated formulation was tested against third-instar le. aegtpti larvae. The aluminum-CMC encapsulated spore-toxin compiex without a photoprotectorwas also affectedby ultraviolet irradiation,resulting in less than 25%olarval mortaiity after inadiation for 24 ius. Results of the larvicidal activity test involving the aluminum-CMC encapsulatedformulations containing the photoprotectors malachite green or ponceau red demonstratedthat major protection of the spore toxin against ultraviolet light was realized, with these particular formulations still causing 55o/oor more larval mortality after24 hrs ofirradiation. In the case ofthe congo red encapsulatedformulation, almost the same larval mortality rates (88%) resulted after irradiation as before (Fig 3)' tn 100 90 80 *Unencapsulated spore-toxin comprex -# *60 Encdpsulation without E50 DhotoDrotector -40 *Enca$sulation with x-30 Malachitegreen 20 +Encapsulation with ponceaured l0 * Encapsulationwith congo red 12 "\A time (h)

FIG.3. - Effrcaciesof B.t.i. spore-toxincomplex microencapsulation formulations after 24 hoursultraviolet inadiation (257 nm) underlaboratory conditions.

DISCUSSION

One of the major drawbackin the use of B.t.i is its rapid inactivationmainly due to sunlight(Priest 1992,Morris 1983,Dunkle and Shasha1989, McGuirre et al. 1996). The inadiationof UV light at wavelengthsranging from 250-380nm hasa detrimentaleffect on the viability and toxicity of B.t (Pusztaiet al. 1991).Yu-Tien et al. (1993)reportedthat.B. thuringiensis.israelensis completely lost its toxicity to mosquitolarvae when exposedto 1.34 x 105 Jlm2 of inadiation at 253 nm. In this regard,our data demonstratedthat UV irradiation,at least under the conditionsthat were used,adversely affected the larvicidal activity of the aluminumCMC-encapsulated formulation of the spore-toxincomplex of B.l.i. containingcongo red as a photoprotectiveagent (Fig 3). Elcin et al. (1995)showed that the aluminum-CMCencapsulation of Bacillus sphaericusdid not protectlarvicidal activity of this agentfrom the detrimentaleffect of UV irradiation.In this study,a rapid lossof toxicity of the unencapsulatedformulation of the 8./.i, spore-toxin for Ae. aegypti larvae was observedwhen exposedto UV inadiationfor up to 24 hrs. Also, the encapsulationof this particularspore-toxin complex without a photoprotectordid little to increasethe toxin's stability and toxicity for mosquitolarvae. Theseresults are consistentwith thoseof other authors.Poszgay et al. (1987) showedthat exposureof B. thuringie,?srsto 40 hrs of UV inadiationresulted in lost activity. Sinceprotein photoinactivationproduced by UV light seemsto be the primary factor affectingthe stability of the toxin, attemptswere madeby othersto achievephotoprotection using exogenous chromophores to absorblight andprotect the delta-endotoxinof this agent. Yu-Tien et al. (1995) achievedphotoprotection of the spore-toxincomplex by addition of melanin.Castro-Franco et al. (1998) tried malachite green,congo red, and charcoalas UV protectorsin formulationsof8. thuringiensisaizuwai. The authorsused the formulationsagainst.s. frugiperda and obtained78, 70 and 45%uv protection, respectively. Ignoffo et al. (1991) showed that activated carbon and polyflavonoidscan also be usedeffectively to protectentomophatic viruses from the adverse effectsof UV irradiation. In this work, the mosteffective chromogen protector at leastup to 24 hrs of UV exposurewurs congo red followed by malachitegreen and ponceauied. However.this photoprotectorsignificantly increased the LCsovalue from 1.074to l.62lg (5l%) and the slopedecreased from 4.2728to 3.4002(20%). castro-Franco et al. (199g)

I4l consideredmalachite green a better photoprotectorthan congo red. This differenceis probablydue to the compositionof theseauthor's formulations compared to ours, As for encapsulationsitself, results indicated that2c/oCMC forms a stableformulation of th'eLr.i. spore-toxincomplex. It was observedthat percentagemosquito larval mortality fo|the 2Yo CMC encapsulatedformulation without photoprotectorcompared to those encapsulated formulationswith ponceaured, malachitegreen, or congored was not statisticallydifferent. The encapsulationusing 2%oaluminum CMC had high capacity to produce consistent capsules.As observedin Fig. l, the decreasein the mosquitomortality after encapsulation was probablydue to the free spore-toxinbeing eliminatedduring filtering and/orwashing. ln the caseof their effect on Ae. aeg)pti larva mortality, there was no statistical preference of which formulationperformed best. Resultsindicate that the encapsulationmaterial and the photo protectorsused are good prospectsas formulatingagents for B.t i. endotoxinto be appliedagainst mosquito larvae. The use of phagostimulantsmight s€rve to increasethe preferencesor feeding rates of mosquitolarvae for theseformulations (Fanar and Ridgway 1994)and, thereby, make these formulationseven more effective.

ACKNOWLEDGMENT

This researchwas supportedby CoSNET.Mexico. Project number 614.99. I acknowledgetechnical assistance of LeopoldoHidalgo Sosaand CelestinoErubiel Prior during project development

LITERATURECITED

Behle, R.W., M.R. McGuirre, and B.S. Shasha.1996. Extending the residualtoxicity of Bacillus thuringiensiswith casein-basedformulations. J. Econ. Entomol. 89: 1399- 1405. Blaustein,L., and J. Margalit. 1991.Indirect effectsof the fairy shrimp,Branchipus schaeferyand two ostracodspecies on Bacillus thuringiensis var. israelensisinduced mortalityin mosquitolarvae. Hydrobiologia. 212:67-73. Castro-Franco,R., J.S. Garcia Alvarado,and L.J. Galan-Wong.1998. An alternative bioinsecticideformulation to encapsulateBacillus thuringiensis 6 toxin and extracts of Agave lechuguillafor the controlof Spodopterafrugiperda Smith.QYTON. 62: 7t-77. de Barjac,H. 1978.A new subspeeiesof Bacillus thuringiensisvery toxic for mosquitoes, Bacillus thuringiensis sero-typeH-14. CompteRendu Academie Sciences Paris SeriesD.286.797-800. Dunkle,R.L., andB.S. Shasha.1989. Response of starchencapsulated Bacillus thuringiensis containinguhaviolet screensto sunlight. Environ.Entomol. l8: 1035-1041. Elcin, Y.M., C Cokmus, and C.S. Sacilik 1995. Aluminum carboximethylcellulose encapsulationof. Bacillus sphaericus2362 for control of Culex spp. (Diptera: Culicidae)larvae. J. Econ.Entomol. 88: 830-834. Ellar, D.J. 1994. Molecular Geneticsof Bacillus thuringiensis6 endotoxin and toxin receptors.Proceed. VI InternationalColloquium on Inv. Path.and Microbiol. Ctrl. Soc.For Inv. Path.Montpellier, France. pp 10-15. Farrar,R.R., and R.L. Ridgway. 1994.Comparative studies of the effectsof nutrientbased phagostimulantson six lepidopterousinsect pests. J. Econ.Entomol. 87: 44-52. Finney,D.I.1977. ProbitAnalysis. Cambridge Univ. Press,Cambridge, United Kingdom. Ignofo,C.M., B.S. Shasha,and M. Shapiro.1991. Sunlight ultraviolet protection of the Heliothis nuclear polyhedrosisvirus through starchencapsulation technology. J.

142 I

Invert.Path. 57:134. Kase, L.E., and L. Branton.1986. Floating particle for improvedcontrol of aquaticinsects. U.S.Pat. 4631857. Leong,K'L., R.J. Cano,and A.M. Kubinski. 1980.Factors affecting Bacillusthuringiensis total fieldpersistence.Environ. Entomol. 9: 593_599. Lord, J'C', andA.H. Undeen.1990. Inhibition of the Bacillusthuringiensis var. israelensis toxin by dissolvedtannins. Environ. Entomol. 19: 1541,_1551. Lowry, o.H., N.J. Rosebrough,A.L. Farr,and R.J. Randall. 195L proteinmeasurement with thefolin phenolreagent. J. Biol. Chem.193: 265_275. McGuirre,M.R., B.s. Shasha,c.E. Eastman,and H.o. sagedchi.1996. starch and fluor- b'asedsprayable formulations: Effect on rainfastnessand solar stability of Bacillus thuringiensis.J. Econ.Entomol.89: g63-g69. Mclaughlin,R.E., H.T. Du.lmage,R.T.L. Alls., D.A. Couch.,I.M. Dame.,R.I.'Hall., M.I. Rose,and P.L.verui. 1984.u.s. standardbioassay for the potenry assessmentof Bacillus thuringiensisvar. israelensisagainst mosquito larvae. Buli. Entomol.Soc. Amer.pp.26-29. Morris, O.N. 1983. Protgcli^ol-o,lBacillus thuringiensisfrom inactivationfrom sunlight. Can.Entomol. ll5: 1215-1227. Ohana,B., J. Margalit andZ. Barak. 1987.Fate of Bacillusthuringiensis subsp lsraelensrs undersimulated field conditions.Appl. Environ.Microbiol. 5r' sza-a:t. Pearson, D., and O.P.Ward.1988. Effect of cultureconditions on growth andsporulation of Bacillus thuringiensis subsp.israelensls and developmento-f media for iroduction of theprotein crystal endotoxin. Biotechnol. Lett 20:451456. Pozsgay,M', P. Fast.,H. Kaplan,and p.R. carcy. r9g7.The effectof sunlighton the protein crystalsftom Bacillus thuringiensis var. kurstaki HDI and NRotz: A raman spectoscopystudy. J. Invert.path. 50:620_622. Priest, F.G. 1992.A review.Biological control of mosquitoesand otherbiting flies by Bacillussphaericw a',,dBacillus thuringiensis. J. Appl. Bacteriol.72:357-369. Pusztai, M., P. Fast, L. G-ringorten,H. Kaplan, T. Lessard,and p. R. carey. 1991.The mechanismof sunlight-mediatedinactivation of Bacillus thuringiensiscrystals. J. Biochem.273:4347. Ridgway,R.L., v.L. Illum, R.R. Farrar,D.D. calvin, S.J.Fleischer, and M.N. Inscoe.1996. Granularmatrix formulation of Bacillus thuringiensis for control of corn borer (Lepidoptera:Pyralidae). J. Econ.Entomol. g9: l0gg_1094. Rushed,S.S., and M.S' Mulla. 1989. Factorsinfluencing ingestion of particulatematerials by mosquitolarvae @iptera: Culicidae). J. Med. Entomol.ZA: ZiO_UA. Tamez-Guerra, P., M.R. McGuine, H. Medrano-Roldan,L.J. Gal6nwong, B.S. shasha,and F.E. Vega. 1996.Sprayable granule formulations for Bacillus thuiingiensis-J.Econ. Entomol.89 : 1424-1430. Tamez-Guerra, P., c. Garcia-Guti6nez,H. Medrano-Roldan,L.J. Gal6n wong, and c.F. Sandoval-Coronado.1999. Spray-driedmicroencapsu lated, B acillus thi,ringiensis formulationsfor the control of Epilachna varvestismulsant. Southwest. Entomol. 24:37-48 Tyanyun,S., and M.S. Mulla. 1999. Field evaluationof new water dispersiblegranular formulations of Bac.illus th.uringiensissvbp. israelensrsand Bairpus sphaericus against culex mosquitoin microcosms.J. of Am. Mosq.control Assoc,15; 356_361. Yu-Tien, L., s' Men-Ju,J- Dar-Der,w. I-wang, c. chin-chi, and c. crre,ng-chen.1993. Protection from ultraviolet irradiation by melanin oi mosquitocid?l activity of Bacillus thuringiensis var. israerensis.J. Invert. p afrt.62: 13 l - li6.

r43 I vol.28 NO.2 SOUTHWESTERNENTOMOLOGIST JI.JN.2OO3

GENETIC VARIABILITY AMONG POPTILATIONSOF TNATOMA LONGIPENNIS. VECTOR OF CHAGASDISEASE IN WESTERNMEXICO

J. A. Martinez-Ibarra2,N. M. B6rcenas-Ortega3,B, Nogueda-Torreso, P. Ramirez-Vallejo3 andM. H. Rodrlguez,L6pez5

Area de EntomologlaM6dica, Laboratoriode SaludPriblica, CentroUniversitario del Sur, Universidadde Guadalajara

ABSTRACT

Enzyme polyrnorphism in Triatoma longipennis Usinger, one of the most important vectors of Chagas diseasein Mexico, was analized using starch gel electrophoresis.Seven geographic locations were sampledin order to determinegene flow among populations and to characterizeintraspecific differences. Of 18 enzymes assayed,three (ES, MDH, and ME) were successfullyresolved and then usedto score geneticvariation. ES was usedto differentiatebetween populations. Both polymorphism and heterozygosityindicated genetic variability in the populations studied. Gene flow betweensome populationswas found to be high. This finding and the low genbtic distancebetween populations indicate similarity among most of the nearby localities, suggestingan important epidemiologicalthreat.

REST]MEN

Se realiz6un estudiopara estimar el polimorfismoenzim6tico entre poblaciones de Triatoma longipennis Usinger, uno de los principales vectores de la enfermedadde Chagasen M6xico, mediantean6lisis electrofor6ticos en gelesde almid6n. Se colectaron ejemplaresen siete localidadescon la finalidad de detectarflujo interpoblacionalde genesy para caracterizardiferencias intraespecificas. De las 18 enzimasprobadas, s6lo tres (ES, MDH y ME) mosharon suficiente resoluci6n. El polimorlismo y la heterocigocidadindicaron variabilidad gen6tica dentro y entrepoblaciones. El flujo de genesentre poblaciones result6 alto. Lo anterior,aunado a la baja distanciagen6tica entre la mayoriade las poblacionescercanas, podria indicar un riesgo epidemiol6gicamente importante.

rHemiptera:Reduviidae 2Areade EntomologlaM6dic4 A. P. 20, 49000Ciudad Guzrn6.n, Jalisco, M6xico. TIREGEP, Colegio de Postgraduados,km. 35.5 CarreteraM6xico-Texcoco, 56230 Montecillo,Estado de M6xico,M6xico. oBecario de COFAA, EscuelaNacional de CienciasBiol6gicas, Instituto Polit6cnico Nacional,Carpio y Plande Ayal4 ColoniaCasco de SantoTom6s, M6xico, D. F. sCentro de InvestigacionesSobre Enfermedades Infecciosas, Instituto Nacional de Salud Priblica,Av. Universidad655, colonia SantaMaria Ahuacatitl6.n,62508 cuemavaca, Morelos,M6xico. t45 INTRODUCTION

Chagasdisease is a public health threat t}roughout l-atin America where it has beenestimated that 16-18million peopleare inGcted (TDR 2000).In Mexico,the blood- suckingbug Triatoma longipennisUsinger (Hemiptera:Reduviidae) is one of the most importantvectors of Trypanosomacruzi Chagas,the causativeagent of Chagasdisease (Velasco-Castrej6n1991, Velasco-Castrej6n et al. 1994,Velasco-Castrej6n and Salazar- Schettino1996, Magall6n-Gast6lumet al. 1998,Vidal-Acosta et al. 2000, Martinez- Ibana et al. 2001).The stateof Jalisco,located in WesternMexico, is consideredone of the Mexican stateswith high risk of transmissionof T. cruzi to humanpopulations by triatomines,specially T. longipennis(Magall6n-Gast6lum et al. 1998,Martinez-Ibana et al. 2001). This statealso had the highestnumber of asymptomaticacute cases detected by parasitoscopictests (Velasco-Cashej6n et a,1.1994). One of the main problems in conholing T. cruzi v@tors is the recolonizationof domiciliary habitatsthrough migration of triatomine bugs from sylvatic to domesticfoci (Wisnivesky-Colli1994, L6pez and Moreno 1995,Noireau et al. 1995, Gajateet al' 1996) and the migration betweenneighboring localities (Dujardin et al. 1996). To determinethe mobility between sylvatic and domiciliary habitats,it is useful to analyze the geneticstructure of representativepopulations and to estimategene flow. Enzyme systemsprovide useful geneticmarkers for population studies,allowing genetic structuresto be elucidated on the basis of polymorphic loci. Isoenzyme polymorphismhas been used to characterizepopulations of T. infestans,the most important vector of T. cruzi in many South American countries. Researchin Bolivia concludedthat the triatomines found inside some homes after a insecticide treatment were survivors and not reinfestantingindividuals (Dujardin et al. 1996). Also in Bolivia, Breniere et al. (1998) investigated the population ge,lreticstructure of this species, concludingthat the panmiticunit was smallerthan previously estimated. In Brazil, Costa et al. (1997) confirmed by isoenzymepolymorphism the existenceof four different chromaticpopulations of Triatoma brasiliensisNeiva. In M6xico, Flores et al. (2000) detected differences between Mexican triatomine species,but failed to find species- specific loci to distinguish among tfuee species of the I phyllosoma complex [I pallidipennis (Stal), L longipennisusinger md T. picturala Usingerl, The presentstudy addressesgenetic variability of three isoenzymes in sylvatic, domiciliary and peridomiciliarypopulations of T. longipenrnis,and the rateof geneflow betweensylvatic anddomiciliary habitats.

MATERI.ALS AND METHODS

Sevenlocations were selectedfor this study.Each location was about30 krn apart from each other, betweenTepic, Nayarit (21" 30' N, 104o54' W) and Guadalajara, Jalisco(20o 40' N, 103o20' W) andbetween Guadalajara and Autl6n, Jalisco (19o 46'N, l04o 22'W). The presenceof triatomines in different habitats in each locality provided appropriateconditions to study gene flow and made feasible the characterizationof geneticvariability betweenpopulations of T. longipennis Two hundred and thirty-eight domiciliary triatomines were collected from the seven localities in the study area, whereasonly I I sylvatic specimenswere collected from a site 10 km North of JaLa,Nayarit (104" 26' N., 2lo 03' w). A total of eight populationswere assayed(Table l). A quarter ofthe thorax from each40-day starvedadult was maceratedseparately in 250 pl of an enzymestabilizer solution (tris-HCl buffer, sacarose,ascorbic acid,

t46 polivinilpinolidona, sodium matasulfate and mercaptoetanol) (Acquaatr 1992)' The extractswere frozen at -30oc until they could be analyzedby electrophoresis. Standard horizontal starch 8el electrophoresis procedures and enzyme developmentconditions describedby Acquaah(1992) were employed.The following 18 (o- *ar-"r were assayed:aconitase (ACO, EC 4.2.1.3);a-glutamate dehy_drogenase cDi{, EC t.4.1.2);aminopeptidase (pEp, EC 3.4.13.9);catalase (cAT, E9 1.11.1.6);6- diaphorase(DIA, EC t.A.i.i); tumarase(FUM, EC 4.2.1.2);glucosephosphate isomerase (CiI, gC S.:.t.e;; glutamateoxalacetate transaminase (GOT' EC 2'6'l'l);. hexokinase leucineaminopeptidase iHf, nC 23.t.1i:iiocitrate dehydrogenase(IDH, EC 1.1.1.42); malicenzyme (ME, EC iiep, Bc z.qJ\.i);malate dehydroginase(MDH, EC 1'1.1.37); i.t.t.+O); mitocondrial isocitrate-NADP(ICDH, EC 1.1.1.42);phosphoglucomutase (pcM, 6c 2.7.5.1);phosphogluconate dehydrogenase (6pGDH, EC 1.1.1.43);soluble esterase(ES, EC j.t.t.t); and superoxidedismutase (SOD, EC l'15'1'l)' The electrophoretic-Gene patterns were recordedphotographically' flow among populations was estimatedwith the equation Fsr = l/(l + 4Nem) (Wrigbt 1931), where Fsr is a measure of gene frequency variance among pop,rtutionr Ld *t" calculated according to Nei (1975); N" representsthe effective populationsize and n, the migration rate. Genetic similarity and distance among populations were estimated with the Identity Index (I) and the GeneticDistance Index (D) of Nei (1975) and the similary ndex (S) of Rogers(1972). Comparisonsof allele frequenciesbetween samples was performedwith ihe independencyG test of Sokat and Rohlf (1979). Values of mean ireterozygositieswere compaf,edwith the Wilcoxon, Mann and Whitney U-test (lnfante- Gil and Zriratede Lata 1997).

RESULTS AND DISCUSSION

Threeout ofthe 18 systemstested gave readable results for the sevenpopulations of T. longipennrs.Enzymes ES, ME andMDH wereresolved and showed only onelocus each.FUM, GPI, HK, LAP, and PEP showedno activity, ACO, cr-GDH,CAT, DIA, GOT, ICDH, IDH, 6PGDH, PGM and SOD producedonly weak bands and were discardedfrom the analysis. Polymorphismwas observedonly in ES; ME and MDH were monomorphic.A locus was consideredpolymorphic when the frequencyof the most common allele was not higher than99% in at least one of the samples.The polymorphic isozymeprofiles are shownin Table l. The allele frequencieswere different amongsome populations; Sylvatic Jala, San Martin Hidalgo and Autlin populationswere the most different. Alleles for the ES locus in SylvaticJala were sigrificantly different(P < 0.05)rN were allelesfor the ES locusin SanMartin Hidalgoand Autl6n @ < 0'05) (Tablel)' All Fsr values*ere zero and N.m was 0.25,Nei's geneticdistance (D) ranged between0.001 and 0.038. Gene flow was 0.25, indicating low migration between betweenthe -syivaticpopulations. Nei's geneticdistance (D) betweenpopulations was low,.except Jala population and the other four populations from Nayarit (even the nearest Jala population, l0 km distant) and c6rdenas population. In agreementwith the low ge,neticdistance, similarity (S) was substantialin most populations and low betweenthe s/vatic Jala population and the other four populations from the Nayarit state, and between san Martin Hidalgo and the neaf,est population, c4rdenas (Table 2). Heterozygosity was similar (zero) in all populations since heterozygoteswere not detected.

r47 TABLE 1. Allele Frequencies polymorphic of Locus @S) in populationsgfz ls from WesternMexico. Population _- Allele t00 90

Puga" 4l 0 0.93 0.07 Xalisco" 34 0 0.94 0.06 Compostela" 43 0.02 0.91 0.07 Sylvatic Jalab l0 0.30 0.70 0 Jdd ll 0 1.0 0 SanMartln Hidalgo. 63 0.1I 0.8r 0.08 Crirdenas" 33 0 1.0 0 Autl6n" t2 0.83 0 nh" Groups of locations according to their

TABLE 2. Nei's GeneticDistances @) (Above Diagonal)and Roger's Similarity (s) BelowDiagonal) between Populations of T. lonsipennis. Population Puga Xalisco comp. S-Jara ldaffi AutIdn Martin _ __.. Hidalgo Puga 0.000 0.000 0.019 0.001 0.002 0.001 0.000 Xalisco 0.833 0.000 0.023 0.000 0.002 0.000 0.001 Compostela0.325 o,+:o 0.027 0.001 0.001 0.001 0.002 S-Jala 0.084 0.1l0 0.093 0.038 0.008 0.038 0.001 Jala 0.382 0.369 0.733 0.080 0.005 0.000 0.005 Sn.Martin 0.154 0.138 0.224 0.075 0.219 0.005 0.001 Hidalgo Cardenas 0.282 0.157 0.296 0.056 0.609 0.071 0.005 Autlan 0.405 0.051 0.325 0.830 0.209 0]67 0.063

Differencesin allelic frequenciesbetween populations allowed populations to be divided into threegroups: (a) SylvaticJala, (b) SanMartin Hidalgo and Autlan and (c) the remaining four localities. Since sylvatic Jala was the only sylvatic population collected,it was the most different population,as expected.San Martln Hidalgo and Autlan populationswere similar to eachother, but differentfrom the otherpopulations, perhapsbecause tley were the two only locationswhere most specimenswere collected in or immediatelyoutside of sleepingrooms, which could be associatedwith a higher rate of preferencefor people as a blood meal source. on the other hand, the population of C6rdenaswas different from that of San Martin Hidalgo in spite of the fact that they are locatedno more than l0 km from eachother. The differencein this casemay be due to the habitats(sleeping rooms and chicken roosts in San Martin Hidalgo and backyardsin c6rdenas)where triatomines were collected. since most houses in the latter are surroundedby brushwood,an active colonization by triatomines from the brushwoodto the close backyardscould be inferred, different from the more stablepopulations in San Martin Hidalgo. Unforhrnately, no sylvatic specimenscould be collected from the brushwoodto corroboratesuch speculation.

148 Heterozygosity was zero in all populations, since only homozygotes were detected.Accoiding io Pasteuret al, (1987),the lack ofheterozygotesmay de explalned by undetectednull alleles. This explanationwas discardedbecause an absenceof izymatic activity (homozygotesfor the null allele) was not observedin any individual' Two other explanaiionsinciude unknownlocally actingselective factors.and biological factors impe,rling random mating, Selection against heterozygotes is uncommorq from especially io. .iry*. loci, and non-random matings produce local -departures alleles allowing the recognitionof 'nrop*-i*lt. Howevei in the absenceof fixed diagrrostic sets of different genotypes,this hypothesisis difficult to refute. Genetic similuity amongdistant localities doesnot favor this hypothesis' The subdivision of one locality into smallerpanmictic subunits(i.e., the walhund subunits effect) appearsto be the most likely explanationfor the overall results. If the that aiff#O in allele frequencyat a locus with two alleles, and a sampleset is collected includessome members from eachunit, then a deficiencyin the numberof heterozygotes will be observedrelative to predictions under Hardy-Weinbergequilibrium conditions 1928).The ,.*on, why thesesubunits differed in genefrequencies could be lWatrtunO of nil or low levels of genetic exchange,allowing genetic drift, and probable founding (low popJutiott. by very flw females.tnlow or recenthouse infestations by triatomines may arise from a single in[".i"ityl, it is iossiUte that a population at a single site 1998)' female,'Accor[ingassuggestea fot T. infestansin Bolivia @reniereet al' to our resrilts, the sevendomiciliary populationswere very similar and It could be could be consideredpart of a larger population in the entire study area. and rapid inferred that the obsewed genetic-uttiiormity may be the result of a recent more field Jirp.*A of the speciesin frestem Mexico. rurttrer investigationsinvolving poiulations arenecessary to validatethis hypothesis'

ACKNOWLEDGMENT

Mr. Jos6Mariscal Hem6ndezis thankedfor supportin the collection of triatomines.

LITERATURE CITED Dioscorides Acquaah,G. lgg2. Practical Protein Electrophoresisfor Genetic Research' Press.P. 132. Dujardin' Breniere,S. i., t"t. F. Bosseno,F. Vargas,N. Yaksic,F. Noireau,S. Noel, J. P. and M. Tybayrenc.tgiS. Smailnessof the panmicticunit of Triatomainfestans (Hemiptera:Reduviidae). J' Med. Entomol'35:9Ll-917 ' R. s. costa j., M. G. R. Freitas-sibajev,v. Marchon-silva,M. Quinhones-Pires,and pacheco.1997. Isoenzym.r d"tot variationin populationsof Triatomabrasiliensis (Hemiptera:Reduviidae: Triatominae). Mern. Inst. oswaldo Cruz 92:459464. Dujardin,- -' l. p., l. Cardozo,and C. Schofield. 1996. Genetic analysisof Triatoma lrioto following insecticidal control interventions in central Bolivia. Acta Tropica6l:263-266' B' Flores, A., E. Magall6n-Gast6lum,M. F. Bosseno,R. Ord6flez, F. Lozano-Kasten, ispinoz4 J. Ramsey,and S. F. Breniere.2000. Isoenzymevariability of five pritt ipa-u"tomine vector speciesof chagas diseasein Mexico. Inf. Gen' Evol. 4:I -8. and C' Wisnivesky-Colli' 1996' Gajate,- P. P., M. V. Botlazzi, S. M. Pietrokovsky, Potential colonization of the peridomicile by Triatoma guasayana (Hemiptera: Reduviidae)in Santiagodel Estero,Argentina, J. Med. Entomol.33:635-639.

t49 Infante-Gil,S', andG. P. zirate de Lara.1997. M6todos Estadisticos. Ed. Trillas.p. 643. L6pez, G., and J. Moreno, 1995. Genetic variability and differentiationbetween population of Rhodnius prolitus and R. pallescerrs,vectors of chagas' disease. Mem. hrst.Oswaldo Cruz 90:353-357. Magall6n-Gast6lum,E., N. c. Magdaleno-pefraloza,G. Katthain-Duchateau,F. Trujillo- contreras,F. J. Lozano-Kasten,and R. Hemandez-Guti6rrez.199g. Distibuci6n de los vectoresde la enfe-rmedadde chagas (Hemiptera:Reduviidae: Triatominae), en el estadode Jalisco, M6xico. Rev. Biomed. 9:l5l_157. Martinez-Ibarra,J. A., ry.. M. Barcenas-Ortega,B. Nogueda_Torres,R. Alejandre_ Aguilar' M. L. Rodriguez,E. Magail6n-Gast€lum,-y.Llpez-Martlnez, and J. Romero-N6poles.2001. Role of two Triatoma (Hemiptera: Reduviidae: Triatominae)species in the transmissionof rrypanosomacruzi (KinetopLastida: Trypanosomatidae) to man in the westcoast of Mexico. Mem. Inst. oswaldo cruz 96:14l-144. Nei' M. 1975.Molecular Population Genetics and Evolution. North Holland,Amsterdam. Noireau, F,, M, F. Bosseno,R. CarrasconJ. Telleria,F, Vargas,C. Camacho,N. yaksic, and S. F. Breniere.1995. sylvatic triatomines(Hemipiera: Reduviidaej in Bolivia: trends toward domesticity and possible infection with Trypanosoma crazi (Kinetoplastida:T4panosomatidae). J. Med. Entomol.32:59a_56g. Pasteur, N., J. Pasteur,F. Bonhomme,J. catalan et J. Britton Davidian. 19g7.Manuel Technique de G6netique par Electrophoresedes prot6ines: Techniqueset Documentation,Lavoisier, paris. Rogers,J' s. 1972.Measures of geneticsimilarity and geneticdistance. Univ. Tex. publ. 7213:145-153. Sokal,R. R., andF. J. Rohlf. 1979.Biometria. H. Blume,Madris. Tropical Disease Research (TDR). 2000. Intervention research on chagas Disease special Programme for Researchand Training in Tropical Diseases.lr.o.n.;. or the UNDP/ World Bank/wHo. Velasco-Castrej6n,O. 1991.La Enfermedadde Chagas.Publicaci6n Tecnica del Instituto de Diagn6sticoy ReferenciaEpidemiol6gica CnDnnl, p. 75. M6xico. velasco-cashej6n,o., c. Guzm6nBracho, and S. Ib6fiez-Bemal.1994. Enfermedad de chagas' pp' 279-292. ln J. L. valdespino-G6mez,o. Velasco-castrej6n,A. Escobar-Guti6rrez,A. del Rio-znlezzi, S. Ibafiez-Bernal, and c. Magos-L6pez (eds.).Enfermedades Tropicales en M€xico. Instituto Nacional de Diagn6sticby ReferenciaEpidemiol6gicos, Secretaria de Salud.M6xico. p. velasco-castej6n,o., and M. salazar-schettino.1996. Enfermedadde chagas en M6xico, pp. 28-29. In C. J. Schofreld, J. p. Dujardin, and J. Jurberg(eds). Proceedingsof the International Workshop on Population Geneticsand Connol of Triatominae,Santo Domingo de los colorados,Ecuador. INDRE, Mexico city, l16 pp. vidal-Acosta, v, Ib6ffez-BemalS, and Martinez-camposc. 2000. Infecci6n natural de chinchesTriatominae con Trypanosomaentzi asociadasa la viviendahumana en M€xico.Sal. Prib. M6x. 42t496-503. Wahlund S. 1928.Zusammensetzung von populationenund korrelation-erscheinungen standpunktder vererbungslehreausbetrachtet. Hereditas I I :65_106. wisnivesky-colli, c. 1994.Triatominos vectores secundarios de Trypanosomacruzi; su domiciliaci6npotencial. Talleres 3:83-89. Wright, S. I 931 . Evolutionin mendelianpopulations. Genetics 16:97 _l 59.

150 vol.28 NO,2 SOUTHWESTERNENTOMOLOGIST JUN. 2003 SCIENTIFICNOTE

OCCURENCE OF THE MAIZE BILLBUG, SPHENOPHORUSAAIDIS IN EASTERNGAMAGRASS'

D. L. Maas',T. L. Springef,andD. C. Amolda

Easterngamagrass, Tripsacum dactyloides (L.) L., has a potentialto produce21,300 kg/ha(19,000 lbs/aoe) per yearofcured hay (Barnett1988) or an estimated78'll2 kglha (70-100 lbs/acre)of pwe live seed(Flaim 2001). Maize billbug, Sphenophorus maidis (C\rttenden), is a well documentedpest of com, Zea mays L', as well as sorghum, Sorghum spp., cattails, Typha spp.,and many speciesof wild grasses,reeds, rushes and sedges(Hayes 1920, Fenton 1940, Vawie 1951, Manley 1999). Satterttwait (1919) reported the maize billbug utilized com and eastem gamagr.rssas food sources,but no observationswere madeof damageto easterngamagrass stands. Damageto the previous year's reproductiveand vegetativeculms of eastemgamagrass was noted in April 2002 in plantings at Woodward, Oklahoma,in Woodward County and near Fort Supply, Oklahoma, in Harper County. Undergroundshoots were hollowed out during larval feeding (Fig. l) and then utilized for pupation.

FIG. l. Larval stage of maize billbug, Sphenophorusmaidis (ChrftendeQ,consuming the underground stalk of eastem gurmagrass,Tripsacum dactyloides (L.) L. Inset is larvae removedfrom stalk.Scale: bar equals5mm.

'Coleoptera:Curculionidae 'Contribution of the USDA-ARS SouthemPlains RangeRes. Stn., Woodward, Oklahoma and the Oklahoma Agric. Exp. Stn., Stillwater, Oklahoma. All programsand servicesof the USDA are offered on a nondiscriminatorybasis without regard to race, color, national origin,religion, sex, age, marit4l status, or handicap. 3USDA-ARS-SPRRS,2000 I 8s Street,Woodward, Oklahoma 73 80 1. 4Departmentof Entomology,Oklahoma State University, Stillwater, Oklahoma 74078. Acceptedfor publication8 February2003. Larval feeding also damagednearby shoots either by direct damageor by allowing accessto opportunistic organisms. Injury to shoot basesby newly hatched larvaq often resultedin partial diebackofone side ofthe leafblade from the midrib outwards. characteristic adult billbug feeding injury in com (Satterthwait l9l9), consisiing of transverserows of holes acrossthe blade,was observedin easterngamagrass plants. Fifty plants randomly selected from 800 plants in a six-year-old easterngamagrass germplasmnursery at Woodward, Oklahoma,were found to be 100 % infested. Similar infestationswere found in a six-year-oldseed productionblock of'FGT-I' eastem garnagrassgermplasm (Dewald and Kindiger 1996)at woodward anda l2-yearold grazed 'Pete' easterngamagrass (Fine et al. 1990)pasture near Fort Supply. Adult, larval and pupal stages of the maize billbug were collected from infested plants. Taxonomic identification of adult maize billbugs was confirmed by Don Arnold, Entomologist, oklahoma State university, stillwater, oklahoma. voucher specimenswere depositedat the K. C. Emerson Entomological Museum, Oklahoma State University, Stillwater, Oklahoma. Damageinflicted dwing the life cycle of the pest will have a negativeeconomic impact on seedproduction from the loss of reproductivetillers. Maize billbug damagemay also contibute to the rate of center'die out' of the plant crown of easterngamagrass. Although eastem gamagrassis a perennialrelative of com, the control mea$uesused in com production may not be effective for gamagrass.Research is neededto characterizethe life cycle of maize billbug in easterngamagrass and to determineeffective methodsof control.

LITERATURE CITED Bamett, F. 1988. Introduction to eastem gamagrass. Proceedingsof the Eastem GamagrassWorkshop, Manhattan, KS. 4 pp. Dewald, C.L., and B. Kindiger. 1996. Registrationof FGT-I eastem gamagrass germplasm.Crop Sci. 36: 219. Fenton,F.A., 1940. The insectpest record for Oklatroma- 1939.Proc. Okla. Acad. Sci. 20:99-102. Fine, G.L., F.L. Barnett, K.L. Anderson,R.D. Lippert, and E.T, Jacobson. 1990. 'Pete' Registrationof eastemgamagrass. Crop Sci. 30:741-742. Flaim, J.2001. Dan and Jan Shepherd,Shepherd Farms, Clifton Hill, Missouri. New AmericanFarmer. 41 -43. Hayes,W. 1920. The maizebillbug or elephantbug(Sphenophorus maidis Chittl). Tech. Bull. 6. KansasAgric. Exp. Stn. Manhattan,KS. 4l pp. Manley, D. 1999. Billbugs on com. Entomologylnsect Information Series.Clemson Univ. Coop.Ext. Bull. Clemson,SC. I pp. Satterthwait,A.F. 1919. How to control billbugs: Destructiveto cereal and forage crops. USDA FarmersBull. 1003.Washington, DC. 23 pp. Vawie, P. 1951. Revisionof the genusCalendra (formerly Sphenophorus) in the United Statesand Mexico. (Coleopter4Curculionidae). Bull. Am. Mus. Nat. Hist. 98: 35-l85.

152 vol.29 NO.2 SOUTHWESTBRNENTOMOLOGIST JUN.2003 SCIENTIFICNOTE

PRESENTSTATUS OF TRICHOBAMSCHAMPIONI BARBER (COLEOPTERA:CURCULIONIDAE)IN MEXICO

RobertoAntonio Huerta Paniagua,Ndstor BautistaMartlnezy JesrisRomero N6poles Institutode Fitosanidad,Colegio de Postgraduados.Km. 36.5,Carretera Mexico- Texcoco.Montecillo, Texcoco, Estado de M6xico.CP 56230

Approximately 45,043hectares of husk tomato,Physalis ixocarpa Brot., resulting in the productionof >580,000tons with an estimatedvalue of ca. US $176 million, are cultivated annually in Mexico. The main producer-statesare Sinaloa,Jalisco, Michoac6n, Puebla, and the state of Mexico (SAGARPA 2002). The most important insect pests attacking husk tomato are Trialeuurodes vaporarioinrrn Westwood, Lema trtlineata daturaphila Kogan and Goeden, Epitrix sp., Macrodactylus mexicanus Bttrm., Trichobaris rnucorea (LeConte ), Trichobaris sp., Heliothis subflexa (Gnenfle),Myzus persicae (Sulzer), Diabrotica undecimpunctata Huold, Euschistus sp,, Murgantia histrionica (Hahn), Lygas sp., Draeculacephala sp., Paratrioza cockerelli (Sulc), and Meromyzasp. (Jim6nezet al. 1992,Gorzillez 1994). Collectionsof insectson husktomato crops, conducted in early2000, in the states of Mexico, Morelos,and Fueblaindicate an altitudinaldistribution of at leasttwo pests that attack husk tomato: Heliothis subflexa Guen6eand Trichobaris championi Barber. H. subflexa is found in denserpopulations at altitudes equal to or higher than 1,800 m abovesea level, whereit is the major pest ofthe crop; the presenceof T championiis negligible at this altitude.However, between 1,800 and 2,200 m above sea level, Z ehampioni predominatesas the principal pest of husk tomato, and the incidence of .FI subflexais lessthan 5%. Trichobaris championi (Coleoptera:Curculionidae),commonly known as "borreguillo"(larva) or "capichi" (adult),has recentlybecome the major pest attacking husk tomatoin the vegetableproducing region of the Altiplano Poblano,particularly in the municipalitiesof Acatzingo,Atoyatempan, Cuapiaxtla de Madero,Felipe Angeles, Huixcolotla, Los Reyesde Jirtrez,P6lmar de Bravo, Quecholac,and Tochtepec.This insectwas initially observedas a pest in the mid 1990's;today, damageis considered severe.This insect has reducedproduction from twelve harvestingsoften years ago to three at present, equivalent to a loss of approximately 75% (personal communications with farmersand technicalpersonnel of the StateCommittee of Plant Health of the State ofPuebla). From earlyApril to September2002, twelve sampleswere taken every 15 daysin five different husk tomato plots in the Puebla high plateau. Sampling consistedof selectingfive sitesinside the plot, in which five husk tomatoplants were removedand dissectedto count the insectspresent in the stalks and branches,Additionally, adult insectspresent on the plantsalso were counted.On the basis of thesesamples, it was determinedthat during the vegetative growth stage, from late April to early June, the incidenceofplants infestedby larvaebored into stemsand branches varied between 68.9 and87 .3Y0. At the endof the crop cycle(second week of June), 77.4yo of the plantswere found with late-instar larvae, pupae, and adults inside the stemsdistributed as follows: eggs,0.0olo; larvae boring in stems,1.3%; pupae in cocoons,5l.6Yo; pharate adults inside cocoons,26.6%; and, adults inside stems, 20.5yo. Eight samples taken every 15 days in five different plots in the Puebla high plateaufrom Januaryto April 2002 indicatethat early infestationsby adultsin seedlings

153 occur in the months of March_and April on land bordering empty lots 5 to 7 days after pfalline'_with a percentageof infestationthat variesbetrvien is.r *o sa.s,% 26.8%). livvage one to threeadults were found on the seedlings;typically therewas one, to a lesser degree two, and rarely three. Based on trresJ samptes,it upp.*"d thai the g:t::l@"r of seedlingsinfesred wirh r championi eggs fluctuated b-&ween37.6 and 65.2%. The adultsarrived on the crop,fed on the foliageand stems for three!o five days, mated'and_ began to oviposit one day later on stemsin the middle region of the plani wherethe first ramificationbegins. To oviposit, femalesmade severaloval holes on the stemswith their rosfrum. Holes were approximately3 mm long and l-2 mm deep;the length of the hole was alignedalong the longitudinalaxis of the plant stem.Some of these-holeswere selected by ^femalesfor oviposition.Females deposited eggs on the upper edge,almost on the surface,in the directionofthe lengthofthe oval, insertingonly-one "g-g pq hole.Holes usedfor ovipositionwere differentfrom the othersin that thei trada-smatt "tongue" of stemtissue.covering the eggwhich waswrapped inside a bag.bggs werekidney-Jhaped, yellow, andmeasured 0.5-0.7 mm. During the crop'sseedling stage, adults reo on rotiage, apparentlywithout causingdeath of the leaf or seedling.Later,damaged seedlings tended to recoverso that this damagewas of no economicimportance. _ olefnalry, T. championi was collected from wild plants belonging to the genera Datura, Solanum and Physalis at sites in the stateJ of lguasc-atiintes, c-olima, chihuahua,State of Mexico, Morelos,oaxaca, and veracruz,anJin the FederalDistrict (Mexico city) (Barber1935). At that time,the insectwas reported as a new species,with no mention-of it as a pest.of husk tomato.Thus, this is the first report documentingthe presenceofT. championiin the Pueblahigh plateauand its peststatus on husktomato in this region. We thank Comit6 Estatalde SanidadVegetal del Estadode Pueblafor logistic supportand the Fundaci6nProduce Puebla, A.C. for hnancial supportin the realizationof this work.

LITERATURE CITED

Barber,H.s. 1935.The tobaccoand solamunweevils of the genusTrichobaris. u.s.D.A. Misc.Pub. No.226. Washington, D.C.28 pp. Gotuu6lezA., J. 1994.indice de divirsidad poblacibnaly entomofaunaasociada al cultivo de tomate de c5scara(Physalis *ocarpa Brot.), su fenologia y rendimiento bajo los sistemasde labranzade conservaci6ny convencionalen chapingo,M6xicb. UniversidadAut6noma Chapingo. I l0 pp. Jim6nezG., R., R. DomlnguezR., y A. PeffaL. 1992.plagas insectiles del tomatede c6scara(Physalis ixocarpa Brot.) en Chapingo,M6xico. Rev. Chapingo.16:75- 79. SAGARPA, 2002. Anuario estadisticode la producci6n agricola 2000. Impresi6n electr6nica. centro de Estadistica Agropecuaria. Secretaria de Agricultura, Ganaderia,Desarrollo Rural, Pesca y Alimentaci6n.M6xico, D.F .

lv vol.28 NO.2 SOUTHWESTERNENTOMOLOGIST JUN.2003 SCIENTIFICNOTE

PHYLLOPHAGACRINITA (COLEOPTERA: SCARABAEIDAE): A NEW PESTIN THE riuasrecnsAREA oF MExrco

Luis A. Rodriguez-del-Bosque,Joel Avila-Valdezl, and Enrique Garza-Urbina2

CampoExperimental Rio Bravo, Instituto Nacionalde InvestigacionesForestales, Agricolasy Pecuarias.Apartado Postal 172, Rio Bravo,Tam., M€xico 88900

The white grubPlryllophaga crinita (Burmeister)is an importantsoil insect pest of com, sorghum, wheat, sugarcane, grassland, and furfgrass in Texas and northem Tamaulipas,Mexico (Reinhard 1940, Teetes 1973, Fuchs et al. 1974, Merchantand Crocker 1995, Rodriguez-del-Bosqueet al. 1995). Occasionaldamage has also been reportedin parsley,cabbage, field beans,spinach, and sugarbeet (Reinhard1940, Plapp and Frankie 1976,Rodriguez-del-Bosque 1988). Adults have beenobserved feeding on the foliageof pecans,citnts,Tamarix sp., and sunflowers(Reinhard 1940, Rodriguez-del- Bosque1988), although the host plants range should be numerousduring the flight periods. P. crinita has a one-yearlife cycle in northeasternMexico and southernTexas, although a proportion ofthe population may have a two-year life cycle in northem Texas (Reinhard 1940,Teetes etal.1976, Huftnan and Harding 1980,Rodriguez-del-Bosque et al. 1995). Reproductiveflights of P. crinita in northem Tamaulipaspeak from late April to early June(Rodriguez-del-Bosque et al. 1995);whereas, most larval feeding activity occurs from July to September,decreases gradually from Octoberto December,and ceasesin January- February(Rodrfguez-del-Bosque I 996a). P. crinita is particularlyabundant in northemTamaulipas, where monoculture of grain sorghumfor the last five decadesmight havefavored building up its populationsin this region(Rodriguez-del-Bosque 1984, Rodriguez-del-Bosque et al. 1995).For instance, P. crinita adultscaptured on standardlight trapsin northemTamaulipas (Rodriguez-del- Bosqueet al. 1995) are much higher than those capturedin severallocations of Texas (Reinhard1940, Gaylor and Frankie 1979, Stone 1986). The most southernknown distribution of P. uinita was San Femando,Tamaulipas (Rodriguez-del-Bosqueet d. 1995), 137 km south of the Mexico-USA'border.During mid-September 2000, soil sampling in several grain sorghum fields near Tampico. southemTamaulipas (292 l

' Campo ExperimentalSur de Tamaulipas.Carretera Tampico-Mante km 55, Esraci6nCuauhtCmoc, Tam., Mdxico89610. -a CampoExperimental Ebano. Carretera Valles-Tampico knr 67. ApartadoPostal 87, Ebano,S.L.P., MCxico.

155 "Huastecas" areaof Mexico, nearly500,000 hectares are plantedwith differentfield and horticultural crops.During this period, there have beenlncreasing reports about crops lossesby white grubs in this region, including corn, grain sorghum,soybeans, and onions. To minimize damageby white grubs, chemical conhol is being consideredby growersin this area,where applicationof soil insecticideswas not a commonpractice ii the past. Thesefindings indicate that P. crinita has becomea new potentia-linsect pest in the Huastecasarea of Mexico.

LITERATURE CITED

Fuchs' T. w., J. A. Harding, and r. Dupnik. 1974.New techniquefor evaluationof insecticidesagunst Phyllophaga crinita. J. Econ.Entomol. 67: 455-456. Gaylor.M. J., and G. W. Frankie.1979. The relationshipof rainfall to adultflight activity; and of soil moistureto ovipositionbehavior and egg and first instar survival in Phyllophagacrinita. Environ.Entomol. 8: 591-594. Hufftnan.F. R., and J. A. Harding,1980. Biology of Phyllophagauinita (Btxmeister)in the Lower Rio GrandeValley sugarcane.Southwest. Entomol. 5: 59-64. Merchant, M., E. and R. L. Crocker. 1995. White grubs in Texas turfgrass.Texas AgriculturalExtension Service L- l l3 I , 6 pp. Plapp,F. W., Jr. and G. W. Frankie.1976. Residues of chlorinatedinsecticides in white grubsand soils treatedfor grubcontrol in urbanareas of Texas.Texas Agricultural ExperimentStation Progress Report 3343C,6 pp. Reinhard,H. J. 1940. The life history of Phyllophagalanceolata (Say) and phyllophaga crinita Bwmeister.J. Econ.Entomol. 33 : 572-578. Rodriguez-del-Bosque,L. A. 1984.oviposici6n de PhyllophagauinitaBvrneister sobre diferentescultivos en el nortede Tamaulipas,M6xico. Southwest. Entomol. 9: 184- 186. Rodriguez-del-Bosque,L. A. 1988. Phyllophaga tinita Burmeister (Coleoptera: Melolonthidae):Historia de unaplaga del suelo(1855-1988). Memorias de la III Mesa Redondasobre Plagas del Suelo,Sociedad Mexicana de Entomologia,24 Mayo,Morelia, Michoacrin, Mdxico. pp.53-79. Rodriguez-del-Bosque,L. A. 1996a. Seasonalfeeding by Phyllophagacrinita and Anomala spp. (Coleoptera:Scarabaeidae) larvae in northeasternMexico. J. Entomol.Sci. 3l: 301-305. Rodriguez-del-Bosque,L. A. 1996b.Pupation and adult longevityof Phyllophagacrinita, Anomalaflavipennis and A. foraminosa (Coleoptera:Scarabaeidae). Southwest. Entomol.2l: 55-58. Rodriguez-del-Bosque,L. A., R. L. Crocker, and E. J. Riley. 1995. Diversity and abundanceof Phyllophaga and Anomala speciesin agroecosystemsof northem Tamaulipas,Mexico. Southwest. Entomol. 20:. 55 -59. Stone, J. D. 1986. Time and height of flight of adults of white grubs (Coleoptera: Scarabaeidae)in the southwesternUnited States. Environ. Entomol. 15: 194-197. Teetes,G. L. 1973.Pltyllophaga crinitai damageassessment and control in grainsorghum andwheat. J. Econ.Entomol. 66:773-776. Teetes,G. L., J. Wade,R. C. Mclnyre,and C. A. Schaefer.1976. Distribution and seasonal biology of Plryllophagauinita in the TexasHigh Plains.J. Econ.Entomol. 69: 59- 63.

rfi vol-.28 NO.2 SOUTHWESTERNENTOMOLOGIST JUN.2003

MINUTES OF THE 2OO3ANNUAL MEETING OF THE EXECUTIVE COMMITTEE OF THE SOCIEry OF SOUTHWESTERNENTOMOLOGISTS.

The ExecutiveCommittee met at 3;30p.m., February 24,2003,at the Renaissance Hotel in Oklahoma City during the Annual Meeting of the SouthwesternBranch of the Entomological Society of America. Presentwere PresidentJohn Jackman,Danell Bay, Editor, and Secretary-TreasurerAllen Knutson. The Editor's and Secretary-Treasurer's reports were reviewed and approved. The proposedname change of the Society and progressto compile all back issuesof the SouthwesternEntomologist onto a CD were discussed.The Committeeapproved of thecontract with Bill Samesfor scanningthe back issuesof thejoumal andsupplement and authorized payment of $500.00to Bill Samesfor expenses.The ballots for President-electwere counted, and there being no furtherbusiness, the meetingwas adjournedat 4:00 p.m.

MINUTES OF THE 2OO3ANNUAL MEETING OF THE SOCIETYOF SOUTHWESTERNENTOMOLOGISTS.

The Annual Meeting of the Societywas called to orderby PresidentJohn Jackman at 4:00p.m., February 24,2003, at the RenaissanceHotel, in OklahomaCity, OK duringthe Annual Meeting of the SouthwesternBranch of the EntomologicalSociety of America.The minutesof the 2002 annualmeeting as publishedin the Juneissue of the Southwestem Entomologistwere distributed and approvedas printed. The Secretary-Treasurer'sreport andEditor's report were distributed,reviewed and approved. Editor Bay notedthat only the Supplementwas mailed in Octoberand that the Septemberand Decemberissues were combinedto meett}le US PostOffice requirementthat four issuesbe mailedeach calendar yearending December 31. Appreciationwas extended to CarlosBlanco for assistingseveral membersin Mexico with makingpayment for their membership. PresidentJackrnan, reporting for theElectronic Publication Committee, stated that the TexasA&M web sitehosts some information about the Society,but a standalone web site dedicatedto the Societyis needed.The costto maintaina web site would be about $100-200per month. President Jackman also reported thatBill Sameshad scanned all ofthe back issuesof the SouthwesternEntomologist and Supplementsand that thesewould be made available as a for-sale cD. The needto updatethis cD each year and the sale to libraries was discussed. of a survey of the membershipconducted during the annual membershiprenewal mailing, 65% of the 153 membersresponding said they would be interestedin purchasingthis CD at an estimatedcost of $25.00. ScottRussell and Bart Drees reportedon progressto updatethe Society's membershipbrochure and dishibuted a draft copy. PresidentJackman recognized Pat Morrison andFrank Gilstrap for their serviceon the NominatingCommittee and reported that Jonathan Edelson had been elected President-elect by the majority of the 154 memberswho returnedballots. Also,72%oof the members approvedofthe proposalto changethe name ofthe Societyto the SocietyofSouthwestem Entomologists.President-elect Drees offered to assistthe Society in makingthis transition. PresidentJackman also reminded the members that the Society would sharethe cost with the Branchand provide a one year's membershipto the winnersin the studentposter and StudentPaper Competition. PresidentJackman then thanked the officers of the Societvand

157 its membersfor their supportand passed the gravelto incomingPresident Bart Drees.Editor Bay awardedJohn Jackman with a plaquein recognitionofhis serviceto theSociety. Tfuere beingno otherbusiness a motionto adjournwas made and approved.

Respectfully submitted, Allen Knutson, Secretary-Treasurer

SECRETARY.TREASURER'S REPORT for fi scalyear February l, 2002-January3 1, 2003

Balanceon hand February1,2002 $6,190.21

Income: Membenhips $ 5,660.00 Subscriptions 650.00 PageCharges 27,320.00 Back Issues 247.65 Royalties 65.54

Total Income w941r9 Expenses: Journal: Editor's Fee $ 2,000.00 Printing 25,016.22 SecretaryFee 2,000.00 AdMail Mailing Service 1,317.86 Postage 1,500.00 Supplies 200.00

Society: SecretaryFee $ 1,000.00 Supplies t37.64 Postage 576.98 PresidentPlaque r20.97 Secretary-TreasurerFee 1,500.00 StudentPaper Competition Award 600.00

Total Expenses $35,969.67

Balanceon handJanuary 31,2003 v,163.73

As of January3 I , 2003, therewere 34I memberspaid for 2002 and 96 institutional subscribersin the SouthwesternEntomological Society. There were l0 unpaidpage charges totaling$2,723.00

Respectfully submitted, Allen Knutson, Secretary-Treasurer

158 EDITOR'SREPORT

Therewere 37 manuscripts,totaling 3 14pages, published in the four regularissues of Votume27 of the SouthwestemEntomologist during 2002compared to 42 manuscriptsand 386 pagesin Volume26 for 2001. Volumes3 and4 werecombined into a singleissue in 2002 due to U,S. PostalService mailing constraints.Additionally, SupplementNo. 25, consistingof 137pages, also was published during the year. A total of69 manuscriptswere received for considerationfor publication during 2002,compared to 55 in 2001. A numberof theseare still in thereview process; however, sevenhave been rejected as of this date. The total of seven manuscriptsrejected is a decreaseoffive from the twelve manuscriptsrejected during 200 I , andrepresents a rejection rate of approximatelyl0%.

Editor's FinancialReport

Date Description Receipts/Expenditures Balance 0ll0ll02 BalanceForward $14.94 0UlA02 Postage 2.63 12.31 0U09102 Postage 8.81 3.50 03108102 Postage 9.73 -6.23 04113/02 Postage 9.74 -15.97 Ml27l02 Postage 11.68 -27.65 05113102 From Treasurer 100.00 72.35 06101102 Postage 5.60 66.75 06115102 Postage 15.70 51.05 07113102 Postage 12.42 38.63 08/17102 Postage 41.25 -2.62 09107102 Postage 7.88 -10.50 09118102 From Treasurer 100.00 89.50 lll23l02 Postage 8.43 81.07 12127/02 Postage 15.38 65.69 12130102 Postage 18.12 47,57

CashSummary

BalanceForward 0l l0l 102 $ 14.94 Receipts 200.00 Expenditures t67.37 $ 47.57 EndingBalance

Respectfully submitted, Darrell E. Bay, Editor

AUDIT COMMITTEE REPORT

I haveexamined the financial recordsof the Societyand the Secretary-Treasurer's reportfor February1,2002 though January 31, 2003,and they werefound to be in order.

Respectfullysubmitted Bart Drees

159 vol-.28 NO.3 SOUTHWESTERNENTOMOLOGIST SEPT.2OO3

ECONOMICINJURY LEVEL FORTHE GREENBUG',SCHIZAPHIS GMMINUM,D] OKLAHOMA WINTERWHEAT

S. D. Kindler,N. C. Elliott, K. L. Giles2,and T. A. Royef

USDA-ARS Plant Scienceand Water ConservationResearch Laboratory, 1301 N. Western Road,Stillwater, Oklahoma 74075

ABSTRACT

The effect of greenbug,Schizaphis graminum (Rondani),on the yield of lines of winter wheat Tricicum aestivum L., was studied during four years in central Oklahoma. Each year, 0.4-haof 'Karl' (1993), 'Karl-92' (1995, 1996,and 1997),or '2163' (1997) winter wheat was planted between I and 15 October. One-meter square plots were establishedand infested during the three- to fourleaf growth stage (fall infestations)or during the tillering growth stagein late winter (spring infestations)with varying numbersof biotype-E greenbugs. Infestationsby greenbugsachieved by artificial infestation of plots were great enough to affect the wheat yield from infested plots, but the intensity of infestationsvaried among years and growingseasons. A regressionmodel was constructed to estimateyield loss for the wheatlines testedas a flrnctionof the maximumnumber of greenbugsoccurring per tiller in the plot. The regressionmodel predicteda 14.5kglha (0.22 bu/ac)loss of yield for eachgreenbug per tiller duringyears with nearaverage precipitation levels,and a lossof 34.3kg/ha (0.5 I bu/ac)under severe drought conditions.

INTRODUCTION

We previously reported on a four-year study of the effect of greenbugs,Schizaphis graminurn (Rondani), on yield of winter wheat in the field (Kindler et al. 2002). In that report, a model was presentedthat expressedyield loss of winter wheat as a function of the greenbug-daysaccumulating per tiller during the growing sezrson.Greenbug-days is an intuitively appealingconcept because it relatesyield to the intensityof an infestationby greenbugsby integrating greenbugabundance (number of greenbugsper tiller) during the duration of the infestation, and exhibits a strong relationship to yield of winter wheat infested with greenbugs(Kieckhefer et al. 1994, Kindler et al. 2002). However, the greenbug-dayconcept is not useful as a quantitativemeasure to determineeconomic injury levels in practical integratedpest managementprograms because winter wheat fields are rarely sampledfor greenbugsmore than one or two times per growing season,thus making it impossible to estimateadequately the number of greenbug-daysthat accumulatefor a field during the growing season.

rHomoptera:Aphididae 'Department of Entomologrand Plant Patholory, Oklahoma State University, Stillwater, Oklahoma 74078

r63 The maximumnumber of greenbugsthat occurper tiller in a winter wheatfield is alsoconelated with yield of the wheat. Althoughthe maximumnumber of greenbugsper tiller doesnot predictyield ofwinter wheatas preciselyas greenbug-daysdoes (Kindler et aL 2002), it has the advantageof practical utility. Typically, winter wheat fields in Oklahomaare sampledfor greenbugsonly if a pestmanager has reason to believethat an infestationby greenbugsis building in the field. In this case,the pest mtuurgercan concludethat the maximum numberof greenbugsthat will occur per tiller in the field is the numberobserved during samplingbecause the field will be treatedwith insecticideif this numberexceeds an economicthreshold. Here we constructa linear regressionmodel to relate yield loss causedby greenbugsto the maximum numberof greenbugsthat occursper tiller in a field of winter wheat. In addition,we createa table that lists economicinjury levels basedon the maximum numberof greenbugsper tiller in relation to the cost of insecticidetreatment and the value of the winter wheatgrain.

MATERIALS AND METHODS

A detaileddescription of field experimentalmethods is presentedin Kindler et al. Q002). Therefore,only a brief descriptionis presentedhere. The studywas conducted during four wheatgrowing seasons in a field 7.5 km westof Stillwater,oklahoma. ,Karl' 'Kzrl-92' ,2163, (1993), (1995, 1996,and 1997),or (1997) winter wheatwas plantedin 0.4-haplots between1 and 15 Octobereach year. One-meter-squareplots wereestablished and infested during the three- to fourleaf gro*th stage (fall infestations)or during the tillering growth stagein late winter (spring infestations)with varying numbersof biotype-E greenbugs.Aphids were excludedfrom someplots for the durationof the studyby perioaic treatmentwith malathion@insecticide. Fall infestationswere allowed to persistuntil heavy freeze each (approximately20 December),after which plants were kept free of greenbugs by periodic insecticidetreatment. Plots infestedduring spring were kept aphid free during autumnand were allowedto persistuntil wheatplants headed. Greenbugswere sampledby cutting l0 tillers from randomlocations in eachplot. In the laboratory,the numberof greenbugsper tiller was determinedfor each sample. Sampleswere taken approximately everytwo weeksstarting 7 daysafter initial infestationin fall. Samplingwas discontinued duringthe winter period,when no aphidswere present in any of the plots. when the wheat seed ripened, five 0.3-m samplesof row were harvestedfrom each plot for yield determination. Heterogeneity-of-slopeslinear regression (Neter and Wasserman1977) was used to relateyield for eachplot to the maximumnumber of greenbugsper tiller measuredduring the growing seasonin the plot. Regressionanalyses were accomplishedusing the pRoc REGprocedure ofthe StatisticalAnalysis System (SAS Institute 1990).

RESULTSAND DISCUSSION

Our previous analysisrevealed that slopes of regressionlines relating yield to greenbug-daysdid not differ significantlyfor any ofthe yearsor infestationperiods (fall or spnng) during which wheat was infested,with the exceptionof the 1995-1996growing seasonduring which severedrought occuned in centralOklahoma (Kindler et al. 2002). As would be expected,regression of yield againstthe maximumnumber of greenbugsper tiller was similar to the previously reported model in that slope parametersdid not differ significantlyamong years or infestationperiods except for the 1995-1996wheat growing se€rson.Here we reporttwo regressionmodels, one that includesunique intercepts for each year and wheat line and unique slopes for 1995-96and all other years combined,and a

r64 secondmodel in which a single slope parameteris used to describethe relationship between yield and the maximum n" Gr of greenbugsp"r til.. rl. For the regressionmodel li"ur" with differentslope parameterJfo, th"?rought (r99i:l9tb andgrbwing s€asonswith approximately average or above averagepreiipitation, the coefficient of determinationwas y' = 0.7g, The esiimateof the slopewas equal to -34.3(sE = 5.9)for the growing seasonwith severe drought. The slope estimateio. trr. g.oirrg seasonwith severedrought was over twice in magnitudeof the estimateor_t+.i 1fr: i.[) obtainedfor the other,three growing seasons. The intercepts for various ;-;;g ,.uior* differed *olg- wheat growing selolf for Karl-92 (the only wheat line-tested"in more than one year) from 2,667 kgrha for rg95--1996 4,613 kg/ha in rg97-1g98ir"Lr" rl. As was expected,the -to slopesrelating yield to the maximuir numberor g...obu!, tiller were negative growing fer for all se.rsons.For the regressionmodel withili ro*-gri*1g seasons combined the coefficient of determinationwas 12 = 0.74. The estimalteof ihe slope paxameterwas -17.7 (St = 1.9).

TABLE 1' Parametersfor LinearRegression Models Relating kg/ha Wheat yield (bu/acin Parentheses)to the MaximumNumber of Greenbugsper wheit Tilrer.

GrowingSeason/ Multiple Slopesu SingleSlopeb WheatLine _

199s-r996lKart-92 2,667(39.7) -34.3(-0.s1) 2,268(j3.7) _17.7(_0.26)

1996-1997/ Karl-92 2,880(42.8) -14.s(-0.22) 2,Sgg(43.t) -17.7(_0.26)

1997-1998I Karl-92 4,613(68,6) -14.5(-0.22) 4,687(69.7) _r7.7(_0.26)

1997-1998I 2163 s,494(81.7) _r4.5(_0.22) 5,573(82.s) _17.7(_0,26\

'The linearregression model hastwo slopes,one for growingseasons with approximately p]:.age or aboveaverage precipitation, 'The andone for droughtcondition.. linearregression model has a singleslope parametei for all growingseasons.

_Our_resultsdistinguish an increasein the loss in yield of winter wheatcaused by greenbug infestationsunder conditions of severedrought from that under normal or abovl normalprecipitation. However, our dataare insuflicient to provideguidelines on the nature of the relationshipbetween greenbug infestation levels ani drough*t.Therefore, it seems prudent at this time to use the regressionmodel with a singlJslope parameterfor all growing seasonsin estimatingyield loss of winter wheat cause-cluy gieenbugs. Economic injury levels basedon the yield loss model for all yearsare listed in raUtei in termsof grain valueper bushelin relationto per-acrecontrol costs, Economicinjury levelsrange beiweentwo and 18 greenbugsper tiller dependingon grain value and cfntrol cost. t[e wide rangein economicinjury levelsillustrates theimportance of estimatingthe numberof greenbugs per tiller in the.field before making control decisions,or u"singsequential sampling planswith appropriatelychosen stop-lines (e.g., Giles et ui. zooo; rihen making control decisionsfor greenbugsin winter wheatin Oklahoma.

165 TABLE 2. Economictnjury Levelsfor Greenbugs(Number /Tiller) in Relationto Control Costand Grain Value.

GrainValue ($ / bu) ControlCost ($ / ac) 4.00 5.00 6.00 7.00 8.00 9.00 10.00 11.00 12.00 2.50 6 8 9 lt la t4 t5 17 l8 3.00 5 6 8 9 l0 t2 l3 t4 15 3.50 4 5 7 8 9 10 ll 12 13 4.00 4 5 6 7 8 9 10 1l t2

4.50 J 4 5 6 8 9 9 l0

5.00 J 4 5 5 6 8 89

).f u J J 4 5 6 6 7 88

6.00 J 4 + ) 6 6 78

A 6.50 2 J 5 5 6 77

A 7.00 2 J J 4 5 5 67

ACKNOWLEDGMENT

We are grateful to Monte Anderson,Melissa Burrows, Wade French,Kane Jackson, Tim Johnson,Keith Mirkes, Perry Shelby,and Justin Spurlin for technical assistancewith this project. Louis Heslerand Leroy Brooksprovided helpfi.rl reviews of the initial version of the manuscript.Mention of tradenames or commercialproducts in this articleis solely for the purposeof providing specificinformation and doesnot imply recommendationor endorsementby theU.S. Departmentof Agriculture.

LITERATURECITED

Giles, K. L., T. A. Royer,N. C. Elliott, and S. D. Kindler. 2000. Binomial sequential samplingof the greenbugin Oklahomawinter wheat. J. Econ.Entomol. 93: 1522- I 530. Kieckhefer,R. W., N. C. Elliott, W. E. Riedell,and B. W. Fuller. 1994.Yield of spring wheatin relationto level of infestationby greenbugs(Homoptera: Aphididae). Can. Entomol.126:61-66. Kindler, S. D., N. C. Elliott, T. A. Royer,K. L. Giles,F. Tao,and R. Fuentes.2002. Effect of greenbugson winter wheatyield. J. Econ.Entomol. 95 : 89-95. Neter,J., andW. Wasserman.1974. Applied LinearStatistical Models. Richard D. Irwin, Inc.,Homewood, IL. SAS Institute. 1990. SAS/STATUser's Guide, Version 6. FourthEdition. SAS Institute. Cary,NC.

r66 vol.28 NO.3 SOUTHWESTERNENTOMOLOGIST SEPT.2OO3

OUPOSITION BY LYGUS HESPERUSIAI{D ITS EGGPARASITOID, ANAPHESIOIB, TN COTTON, ALFALFA AI.ID A WILD MUSTARD

CharlesG. Jackson USDA, ARS, PWA WesternCotton ResearchLaboratory, 4135E. BroadwayRd., phoenix,fuizona 85040-9903

ABSTRACT

oviposition patterns of Lygus hesperusKnight and its parasitoid Anryhes iole Girault were observedin three host plants. The majority of the z. hesperuseggs were depositedin the upper third of cotton" Gossltpiumhirstum L, Var. Deltapine lO, ana mixed cultivars of alfalf4 Medicago sativa L, and in the uppt 213of wild mustard, Londgn rccke/.,Sisymbrtum irio L., plantsin cages. In choicetests, Z. iesperns deposited significantly more eggs (787o) in alfalfa than in cotton but did not show a significant preferencebetween alfalfa and London rocka. Anqhes iole readrilyparasitized eggs of L. hesperusin all three plant types, but appearedto searchurd parasitize egg" on all sectionsofthe plants rather than restricl searchingto the areaswhere most ofthe host €ggswer€ located. Significantly, more z. hesperuseggs were parasitizedin alfalfa than in cottoq byt a greater percentageof eggswere parasitizedin cotton. More eggs were parasitized in London rocket than in alfalfa, but the p€rcentagesparasitized w€re not significantly different.

INTRODUCTION

Lygus hespernsKnight is the predominatespecies ofthreerlgus spp. that feed on agncultural crops in the southwesternu.s.A (Grahamet al. 1982).-L. neiperus feedson a wide variety gf host plants (scott 1977)nd migratesfrom one host to anotheras they becomeavailable during the year (Graham et al. 1986). Alfalfa -lu{edicagolyertivwL., and a wild mustard,London rocket,sisymbrium irio L., are commonhost piantsof Lygus speciesin southernArizona crop areas(stitt 1949;Fye 1972,l9z5; Graham* at. tieo;. cotto4 Gossltpiun hirsulum L., is not a preferredhost of Z. leryerus (Sevacherianand Stern 1974),but cotton may suffer damagewhen large numbersof L. hespentsmove from recentlycut alfalfato feed. Strip-cuttingalfalfa and interptanting alfalfa with coton were proposedas methodsto preventthis movement(Stem et al. 1967,stern 1969). In additioq populations of L. hesperusmay develop on London rocket and then move to alfalfa when the weedsmature (Grahamet al. 1986). Damageto alfalfa grown for hay is not consideredserious; howo,req damagecan be economicaily severewhen alfalfa is grown for seed. London rocket germinatesin late fall and flowers from Decemberto April (Graham et d. 1986), but the growth period may be extendedor contractedby the timing and amountof rainfall @arker 1972). Standsare usually restrictedto paiches along roadsides,ditch banks,and ends offields but may cover large areasoffallow

I Heteroptera: Miridae ' Hymenoptera: Mymaridae

r67 fields. London rocket also grows as a weed in alfalfa during winter and early spring, when alfalfa is still growing slowly so that the two plant types may be found growing in the sameagricultural areaor interspersedin the samefield. The mymarid wasp,Anqhes iole Ctta;lrt" is a parasitoidof Lygus spp. eggs and the most common parasitoid of L. hesperus(Graham and Jackson 1982, Jacksonand Graham 1983, Graham et al. 1986). A. iole has been reared from alfalfa and London rocket $ems and from cotton leavescollected from the field (Grahamand Jackson1982, Graham et al. 1986). tr162s spp. deposit large numbersof eggs in alfalfa and London rocket in southemArizona and large perc€ntag€softhese eg$ may be parasitizedbyl. iole s thtt both plant types are key hosts for survival and population increaseof both pest and parasitoid. The relationship benrreenLygus spp. and its hostg alfalfa and London rocket, and the effect of this relationship on parasitism of Lygas spp. are of obviousimportance. An understandingofthis relationshipis necessarybefore a deoision to corfirol London rocket in alfalfa fields and in the zurroundingareas is made. I did a seriesof teststo determineovipositional preference by L. heryerustnd A. iole for plant type (cotton vs alfalfa and alfalfa vs London rocka) and to further determineegg distributionon eachofthese plant types.

MATERIALS A}.ID METHODS

Cotton (DPL 16), alfalfa (mixed cultivars), and London rocket plants were used for experimentson ovipositional behavior of A, iole ud L. lreryerus. The stagesof the plantsthat are most susceptibleto damageor those on which the largestpopulations of Z. heqerus occur were used. For alfalf4 thesewere plants with buds or blooms and for cotton"these were plants with small buds (squares)or blooms. Plant stagewas less critical for London rocket becauseblooms are presentduring much ofits grounh. For eachset of testq plantsof about the sameheight were used. All plants w€re grown in the greenhouseto prevent oviposition by wild L. heryerus,with the exceptionof some London rocket plants which were transplantedfrom the field. These were held in a greenhouseuntil any Zygus spp. eggshad hatchedand parasitoidshad emerged. The L. hesperuswere collected in the field from alfalfa or London rocket with the o

168 1995). The numbers of L. hesperuseggs and parasitoidsreared from the stems were lransformed by square root and percentageswere transformed by arcsin square root. Numbersare given as the meanand standarderror of the mean. Oviposition by L. hespentson Cotton qrd Poasitisn by A. iole. Eight Type,l cageswere usedto determineZ. hesperusovipositional patternson @tton and parasilism of L. heryera-seggs by l. iole. Potted cotton plants were placedin the cages,rwo ptants per pot and four pots per cage. Twenty L. lvsperus femalesw€re put into eachcage -Afterfor 48 h, then removed and replacedwith eight on+day-old A. iole femaresfor 24 h. the unparasitizedL. hesperuseggs on the plants hatched(8-10 days),the plants w€re qrt into l5-cm sectionsfrom the apex and the numbersofhatched and unhatchedeggs in eachsection countedto determinethe numberof eggsat eachheight. Atl sectionsolthe two plants in each pot were held togetherto obtain oviposition data on A. iole so that no information on height preferencewas collected for the parasitoid. Ten additional days were allowed for parasitoid emergence,after which the stemswere examinedfor eggs with the caps (opercula) still intact. Thesewere dissectedfor evidenceof developmd-g but unemergedparasitoids were addedto the numberemerged to obtain a total count of parasitoids. plants - Porasitism of L. hesperusEggs at Different Heights on cotton by A. iole. Individual cotton plants were enclosedwith sectional cages(Type2) to daermine ifz{. iole prefened a particular height on cotton plants to parasitizeL. hesperuseggs. The numberofleaves in each l5-cm seclionwas countedbefore the cageswere put into plac€. one L. hespenrsfemale per large leaf was confined to eachsection for 48 11after which the sectionalcages were removed,the plantsplaced individually in Type.l cages,and the whole plant exposedto femaleparasitoids for 24 h. One femaleparasitoid for eachfive leaves(rounded to the nearestfive) was releasedinto eachofnine cageswith plants with two sectionsas was one femalefor each2.5 leavesin six cageswith plantswith three sections. One week after exposureto the parasitoids,the plant stemswere dividedinto the marked l5-cm sections,the numbersof eggs were counted, and each s€ction was held separatelyuntil parasitoidemergence. The plantssections were re-examinedg-10 days later for eggs containing unemergedparasitoidq and the total number of emergedand unemergedparasitoids was recorded. ovipositiut by L. hesperusqd Parasitism by A. iole on cotton vs.Atfafa. Five Type-l cages,each containing two pots with nro cotton plants each and two pots with one alfalfa plant eaclq were set up to test preferencebetween these two plant types. All but six of the longer stemswere clipped from eachalfalfa plant, so that tlrere were 12 alfalfa stemsand four cotton plants in eachcage. The pots with the two plant types were placedin the four cornersaltemately around the cageand the positionswere rotatedin the next cage. Alfalfa plants with buds, but not blooms, and cotton plants with flower budswere used. The plantsin eachcage were exposed to 20 femalez . hesperusfor 4g h andthen to 16A. iole females(one per cottonplant and alfalfa stem) for 24h. oviposition by L. hesperusand Parasitism by A. iote on Alfatfa vs. London RfPt Type-l cageswere usedto determinethe preferenceby L. hespenrstooviposit in alfalfa or London rocket and the preferenceby A. iote to parasitizeL. neryerus iggs in thesetwo plant species. The number of eggslaid by L. lrcwrus in the two planitypes was testedin 19 replicate cages. Z. hesperus(20 per cage)were enclosedwith-trno aliirlfa plants and two London rocket plants in each cage for three days, after which they were removedand replacedwith eight u4.iole for tfuer',days, so that both pestsand parasitoids weregiven a choiceof plants. Plantswere selected for similarheights. The number and percentageof Z. hesperuseggs parasitized by A. iole werie determinedin the same19 cagesplus eight additionalcages (27 repticatestotal). For the eight additional cages, sixteen potted alfalfa plants and 16 London rocket piants were r69 enclosedindividually (no choice of plant species)with ten r. hesperusfemales for thnee days. The L. hesperuswere removed and the plants were combined in the eight cages. Each cage containedtwo pots with one alfalfa plant in each pot and two pots with one London rocket plant in each. Bight A. iole were releasedin eachcage for three.days. Therefore,only the parasitoidswere given a choiceof the two planttypes in theseeight ca8es. The plantsfrom nine cageswere dividedinto l5-cm sectionsafter exposuretoZ. lwspents andA. iole femalesto determinethe distribution of L. heynts eggs and from four cagesto determinethe distribution ofparasitized eggs. After the eggs hatched,th€ numbersof L. heryents eggsdeposited in each sec{ionwere counted. The plant sections were held an additional ten days or until the adult z{. iole emrcrgedand were counted. The eggs with the caps still intact were examined for unemergedparasitoids. These were countedand added to the numberofemerged parasitoids for a total count.

RESULTSA}ID DISCUSSION

Oviposition by L. hesperuson Cotton and Parasitisn by A. iole. The average heightofthe portionsofcotton plantsexposed to the insectsin the eight cageswas 46.5 cm. There were 105 flower buds (squares)on the plants. L. hesperts deposited54.5o/o of 2,040eggs in the upper l5cm ofthe plants,32.f/o in the sectionl5-30cm belowthe top of the plant, and 12.7%oin the bottom section. This trend for more eggs located in the upp€rpart of plantsis supportedby the work of Benedictet d. (1981)who showedthat 66Yoof L. hesperuseggs were depositedin the top one-third (approximately 24cm) of glandedcotton (Acala 4a2-77); 21Yowere found in the middle third and 13% in the bottomportion. L. hesperusdeposited 8l.4Yo of eggsin the leaves(Table l). Most eggswere depositedin the petioleswithin 2cm ofthe enlargedportion at the baseofthe leafblade. Theseresults agree with thoseofBenedict et al. (1981)who foundthat 69-87VooftheZ. hesperuseggs were depositedin leavesof glanded Ac.ala442-77;64-780/o of thesewere found in the leafpetiolesin the upperhalf ofthe plants. Theeight.,4.ioleineachoftheeightte$cagesparasrtizd42.6+6.91 (17.4o/o\L. heryerus eggswithin a 24-h period. The numberof A. iole rearedfrom L. heryerus eggs did not increasesignificantly with the numberof hosteggs in a plant pair (r:0.2588, P: 0.1526,df = l, 30). The plant sectionsand plant partswere not held separately,so that no datawere obtained on parasitismofeggs in the separatesections. Tingey and Leigh (1974) were concemedthat differencesin plant height might atrwt L. hesperusegg depositionin cottonvarieties of differentheights in cages. They found that L. hesperusoviposited significantly more in the taller plants (varieties)in cageswhen given a choice. The plantsin this test were of similar heightto eliminatea choiceof taller plants. Parasitism of L. hesperasEggs at Diflerent Heights in Cotton Plorts by A. iole. For this test the L. hespentswere confinedto distinct 15-cm sectionsof the plant for oviposition. The parasitoidswere given accessto the entire plant to determineif there was a preferenceto parasitizeL. heryerus eggsin a particular plant seclion. Even though theL. hesperuswere not givena choiceof planthei9ht,72.T/o of the eggswere deposited in the top l5crn of the plantswith two sections(Table 2). In taller plantswith three sectiongsimilar numberswere depositedin the 0-15 (39Yo)and the l5-30-cm (4l.7%o> sections. The percentagesof eggs parasitizedin the different plant sectionswere not significantly different (P > 0.05). Thus, the data show tll.atA. iole did not show a preferencefor searchinga particular heighg but searchedall sectionsofthe plantsused in this test (table 2). Variability in the percentageofeggs parasitizedare partially due to the low numbersof parasitoidsreleased in the cages. Averagesof 2.4 and 4.8 female

170 parasitoidswere releasedin cageswith plantswith two and three sections,respeclively.

TABLE l. Location of L. heqrents Eggs in Various Parts of Cotton Plants Held in

Location eggs traainstem sz 15 lo 0 5.5 Leaf petiole 671 401 153 0 60.0 Leafbase 123 126 52 0 14.8 Leafvein 55 28 5 0 4.3 Leafscar 0 61000.8 Terminalt & 3.1 Axillary branch Stem 61 17 I 0 3.9 Square 14 2 0 0 0.8 Stemof Square 7 J 000.5 Bud" 15 39 27 0 4.0 Leaf petiole T2 26 101.9 Leafbase 4 4 000.4 Leafvein I 0 0 0.05 r Stemswere aivided into l5-cm sectionsfrom the apexof the plants' b For this test, "terminal" is all parts of the plant above the topmost expandedleaf, usually 2-4cm in length. " "Bud" refersto the a,xillarymeristem.

Oviposition by L. hesperusand Parasitism by A. iole on Cotton vs.Alfalfa' Wlrcn offered a choice between alfalfa and cotton plants, L. hesperuspreferred to oviposit in alfalfa (Table 3). Significantlymore (F: 58.7597,df = 1,18,P < 0.0001)eggs were found in alfalfa than in cotton. Overall, 78.2Voof 2,409 L. hespentseggs wer€ deposited in the alfalfa plants. Significantlymore (F = 18.2457,df - 1,18,P < 0.0@4) eggswere parasitizedin alfalfa plants than in cotton plants. However,a larger percentage(F = 5.7521,df = 1,18,P < 0.05) of the eggsin cottonplants were parasitizedthan of thosein alfalfa. Theseresults show thatA. iole searchedboth plant species.Considering that I . rble searchedfor only 24 h, the 22.3Yopuasitism of Z. lrespenrseggs on alfalfa and 36.5Yoon cottonwas substantial. An averageof 5343Yo of the total numberof eggs in both cotton and alfalfa was depositedin the top l5cm andthe numberdecreased with eachsuccessively lower section (Tabb ). This trend was observed for the cotlon plants in previous tests. Approximately one-half of the eggs in atfalfa plants were depositedin the main stem in the top l5cm. Grahamand Jackson(1982) found a similar distributionof Zygzs spp' eggsin field-collected alfalfa stems. In the presentstudy, the different plant sectionswere not put in separategages for parasitoidemergence, so parasitismdata was not obtainedfor the different plant sections of either plant species. However, in a field study by Graham and Jackson(1982), parasitismlevels of 4, 17, 18, and 20%,were found for eggs in successivel5-cm segmentsfrom the top to bottomof alfalfaplants. Lower parasitismlevels of eggsin the

17l o\ sf q t"t +l 2i e \o ;€. -q Lh Ef E \o aFg oi A(rA !\rE + !.1 ;$ *. n\fc t''l A-O 3o I E,I ? I'l; o\N EE e 5lls 00 $ +l *E ; \o 6.9 3 oi EE6 €1,lF9e F $ FI .+ ri E; I @ *' b .t{ + o o\ oq € d€ E' /i N .E€ g s t\ .L' €&o\o= € ru O- ll'lgo^ = rr6 oq \o G' c.; c) {1 + E 8* E o\ 0) r.i oi () €s B trv&t o tr ao .E tl,lF60il! o\ [).c) F. - 9r5 3 fijlc\o : n n U) '5;s t Fllr cl + \o @ g \o 6 EE o 8..5 $ll'$ E * ID il0olE SE ; a Hs * o\ \o 6 Et€ ! El /i il) ac$ (u .E9^e $l: () G) E ill €.Eol H1€ 6 d Esfg 2E Ef;ET 31 oOE;

172 TABLE 3 . L. hesperusOviposition and Parasitismby A. iole in Choice Tests of Cotton andAlfalfa PlantsCaced in a Greenhouse. Plant tlpe No. eggs No. eggs o/oeg83 deposited' parasitized' parasitized" ' *SEM ' +SEM ' *SEM

Cotton 52.6t9.43 17.0+ 2.83 36.5+ 5.06 Sicnificancelevelb <0.0001 0.0004 0.0275 ' Numbers of eggs depositedand parasitizedper plant were transformedto the square root andthe percentageofeggs parasitizedby arcsinsquare root (N = l0). o Significance levels were determinedon plants (alfalfa) or pairs of plants (cotton) by AIIOVA or by Kruskal-Wallaceone-way AI.IOVA on rutks.

TABLE 4. Oviposition by L. hesperusin Sectionsof Cotton and Alfalfa Plants in Choice Testsin in a Greenhouse. Plant sections t 0- 15 33.1+ 5.46 98.9+13.98 s2.5 15-30 t0.7+2.04 53.1+ 6.34 20.3 28.2 30 - 45 7.9+2.89 2l.l + 4.27 15.0 rr.2 > 45 0.9* 0.38 15.2+3.57 1.7 8.1

15 cm in length. b Means are basedon two cotton plants and one alfalfa plant per pot and two pots p€r cagein five oages(N = l0).

tops ofthe plants in the field may havebeen due to rapid plant growtll so that eggsthere would havebeen recently depositedand exposedto the parasitoidsfor a shorterperiod of time. Oviposition by L hesperusand Parasitism by A. iole on AIfufa vs. Inndon Rocleet. The number of eggs depositedby L. heryerusin the two plant specieswas comparedin 19 cages. Therewas no significantdifference (F = 1.6376,df = 1,36,P = 0.2088)in the numberofeggs depositedin alfalfa andLondon rocke plants. Parasitism of L. hesperuseggs in the two plant typeswas comparedby the numberof parasitized eggs and by the percentage of eggs parasitized per cage Q7 ages) (Iable 5)' Significantly fewer parasitizedeggs were found in alfalfa than in London rocka (F = 4.5378, df = 1,53, P = 0.0379), but the percentagesof parasitized €ggs were not significantlydifferent (F = 1.68a7, df =1,52,P = 0.2000)(Table 5). Thesedata indicate that neither L. heryents nor A. iole demonstrateda marlcedpreference for either plant speciesunder the test conditions, but that there was a tendencyto chooseLondon rocket over alfalfa. In this tes! the plants were chosen for the growth stageson which the laxgestLygus spp. populations were found, but differencesin growth stagecould possibly alter preferencefor a particular species. Datafor the distributionofthe eggsof L. hesperusand those parasitized by A. iole in different sectionsofalfalfa and Londonrocket plants were collectedfrom nine cages

173 parasitism IAILE_I Lygus hesperasoviposition and by A. iole in Alfalfa and London Choice No. No. eggs No. eggs Plant aqe Percaqe' N N Alfalfa l9 73.3+7.41 27 4.8+ 0.94 27 5.7+1.17 London rocket l9 91.2r8.47 27 9.8+ t.ZS 27 8.6+ 1.66 P:0.2088 P = 0.0379 P = 0.2000 Num!_erof eggs and numbe.s by Al'IovA of squareroot transformedcounts. The averagenumbers per plant tlpe per cagewere usedto get the I + SEM. o The perce_ntages_o! eggs parasitized were transformed by arcsin square root before analysisby l(ruskal -Wallace one-wayANOVA on ranks.

for the z. hesperuseggs and four cagesfor the parasitizedeggs (Table 6). Mos of the tr. t Pryr esq:^rere deposited.inthe upper30cm in both anla and 6ndon rocke( but therewas a differencein the distributionon the plant species.They largestfroportion of the eggs was found in the top l5cm in alfalfa (5g.7o/o)and in the top-lo"in in rondon (69.70/o). r9gk6. Gratramand Jackson(1982) found similartrends foilygus spp.eggs in alfalfa and London rocket for field-collected stems.

parasitized TABLE 6. Distributionof andunparasitized Eggs of L. heryentsin Alfalfa andLondon Rocket Plants in Cagesin a Greenhouse.

depositedper plantn -frEii[in plant sections" parasltired. Prantsection@ rJFa ffifu- 0 - 15 42.7t4. 18.314. 58.8 26.8 l.l 13.07 - 15 30 18.8r4.2 29.2t6,t4 25.8 42.9 2.r 6.2 30 - 45 8.3X2.26 9.4r 1.84 I1.8 13.8 0.0 0.0 > 45 2.6+ 0.86 tt.9 +.2.30 3.5 16.5 2.6*"nniifu 1.6 the numberand distributionof L. hesperuseggs by sectionon nin" cag", (lg plantsof eachkind). oNumbersare f +SEM.

The data for the distribution of parasitizd L. tresperts eggsare limited, especially for those in alfalfa stems,but a generaltrend for the parasitizedJggs on London rocket is apparentfrom the data (Table 6). Very low levels ofparasitism were touna in the alfalfa stems and no conclusionsare apparent. Graham and Jackson(19g2) studied the distributionofparasitized eggs in field-collectedalfalfa and London iockj stemsover a two-yearperiod. The total numbersof parasitizedeggs in their studywere also low, but trendswere obtained. In alfalfa stems,more parasitized eggs were found in the l5-30-cm section, followed by the 0-15-cm section,but the 0-15-cm section had the lowest percentageofparasitism. Parasitismlevels at 15-30,30-45, and more than 45cm from the apicesof the stemswere all similar (17-202o). In field-collectedLondon rocket stemg the distribution of the number and percentageof parasitizedeggs was more even

174 amongplant sections(Graham and Jackson1982). Percentageparasitism ranged from 2l%ointhel5-30-cm sectionto 45%ointhe 0-15 and 30-45-cm sections. In summary,L. heryerus showeda preferenceto oviposit in alfalfa over cotton in choice tests, but showed no significant preferencebetween alfalfa and Inndon rocket. Anqlws iole parasitizd L. hesperuseggs on all three plant g/pes. The number of eggs par.asitizedwas greatu in alfalfq wherethe larger numberof eggswere deposited,than in cottoq but the percentageofeggs parasitizedwas higher in cotton than alfalfa. There was no difference in the number of eggsdeposited in alfalfa or l.ondon rocket, nor was there a difference in the percentageofeggs parasitized. However, a larger numberofthe eggs in London rocket were parasitized. Lygus hesperustended to oviposit in the top on€-third of cotton plants and the top two-thirds of alfalfa and London rod

ACKNOWLEDGMENT

EugeneNeemann and Michael Plagensassisted with all aspectsof the experiment. Drs. JamesHagler (USDA-ARS,Phoenix, A,Z) andHollis Flint (USDA-ARS,Phoenix, AZ) providedcritical reviewsof an earlierversion of the manuscript.

LITERATURE CITED

Benedict,J. H., T. F. Leigh,J.L.Frazer, and A. H. Hyer. 1981. Ovipositionalbehavior of Lygus heryerus on two cotton g€notypes. Ann. Entomol. Soc. Amer. 74: 392- 394. Fye, R E. 1972. Cotton pest and predatorreservoirs in Awa Valley. Prog. Agric. Ariz. 24: 15-16. Fye, R. E. 1975. Plant host sequenceof major cotton pests in southern Arizona. U.S.D.A, Agric.Res. Serv. Rep. W-24,9p. Graham,H. M., and C. G. Jackson. 1982. Distribution of eggs and parasitesof Lygas spp. (Hemiptera: Miridae), lVaDis spp. (Hemiptera: Nabidae), and Spirsislilzc festims (Say) (Homoptera:Membracidae) on plant stems. Ann. Entomol. Soc. Amer. 75:5G60. Graham,H. M., C. G. Jackson,and G. D. Butler, Jr. 1982. Compositionof the Zygzs complexin somecrop andweed habitats of fuizona. Southwest.Entomol. 7: 105- I 10. Graham,H. M., C. G. Jackson,and J. W. Debolt. 1986.Lygus spp. (Hemiptera: Mridae) and their parasitesin agriculturalareas of southernfuizona. Environ. Entomol. 15:132-142. Jacksoq C. G., and H. M. Graham. 1983. Parasitismof four speciesof Lygus Slemiptera;Miridae) by Amphes ovijenlalus(Hymenoptera: Mymaridae) and an evaluationof otherpossible hosts. Ann. Entomol.Soc. Amer. 76:772-775. Parker, K. F. 1972. An illustratedguide to Arizona weeds. Univ. Arizona Press, Tucson,42.338p. Scott, D. R. 1977. An annotatedlisting of host plants of Zl6zs hesperustfuight. Bull. Entomol.Soc. Amer. 23: 19-22. SevacheriagV., and V. M. Stern. 1974. Itrostplant preferencesof lygus bugs in alfalfa- interplantedcotton fields. Environ.Entomol. 3:761-766. SigmaStat.1995. Version2.0 user'smanual. SPSSSciencg Chicago, IL. Ster4 V. M. 1969. Interplanting alfalfa in cotton to control lygrs bugs and other insect pests. 1lr Komarek, E. V. [ed]. Proc. Tall Timbers Conf. Ecol. Anim. Contr. Habitat Management,No. 1. Tallahassee,FL: Tall TimbersResearch Statio4 p. 55-69.

175 Sterq V. M., R. van den Bosc[ T. F. LeigtL O. D. McCutcheoqW. R. Sallee,C. E. Houstor\and M. J. Garber.l%2. Lyguscontrol by stripcutting alfalfa. univ. California"Agric. Ext. Wer. AXT-24I, if p. stifi' L. L. 1949. Host plant sourcesof Zygz.rspp- infesting the alfalfa seedcop in southernArizona and southeasterncalifbrnia. J. Econ. Eirtomol. 42:93-D. Tingey w. M., and T. F. Leigtt 1974. Height preferencesof lygus bugsfor oviposition on cagedcotton plants. Environ. Entomol. 3: 350_351. zaugg J. L., and M. w..Nierson. 1974. Lvgus hesperus:comparison of orfactory responsebetween laboratory-reared and field collected adults, J. Econ. Entomoi. 67:133-t34.

Punluserlby the U.S.Deparunent of Agrioiture FOROFRCI,AL USE ONLY

176 vol.28 NO.3 SOUTHWESTERNENTOMOLOGIST SEPT.2OO3

INCIDENCE OF THE BEET LEAI'HOPPER. CIRCULIFER TENELLUS (HOMOPTERA:CICADELLIDAE)IN NEW MEXICO CHILE

R. Creamer,J, Carpenter,and J. Rascon

Departmentof Entomology,Plant Pathology, and Weed Science; 'Box 30003,MSC 3BE;New Mexico StateUniversity, Las Cruces,NM 88003

ABSTRACT

Chile fields were sampledfor the beet leafiropper,Circulifer tenellus@aker), vector of beet curly top virus (BCTV) during 2001 and 2002in two New Mexico countiesevery two weeksusing yellow sticky taps. weed hostsof the beet leaflropperand BCTV were collected &om the fields at the sametimes and testedfor virus using PCR. Higher numbers of leafiropperswere collected and more virus detectedin 2001 than in 2002. It both years, leaftropper flights into chile fields initiated in April to May, peaked in June-July, and droppedoff by late october to mid-Decernber.However, the timing of leafhopperflights differedbetween the two countiessampled. virus incidencein the weedswas higher (1.3%) in 2001than2002 (0.06%), and no virus symptomswere observed in theweeds.

INTRODUCTION

Beet curly top virus (BCTV) epidemicshave occurred sporadically in the southem chile pepper growing areas of New Mexico since the first report in the state in 1927 (Crawford 1927). ln tbree of the last l0 years, New Mexico chile sustainedsubstantial lossesdue to curly top. In 2ool, a year with high curly top incidencein New Mexico, the chile yield per acre ww l2o/o less than in 2000 or 2002, years with minimal curly top pressure(NASS-USDA 2003). Chile plants infectedwith curly top are often severely stunted,with chloroticleaves and, if infectedearly, do not producefruit. Chile fruit that ii producedby plants infectedlater in the seasonis usuallysmall and round. No effective control me:iluresare known for curly top on chile. BCTV is a monopartitegeminivirus of the genusCurtovirus, which is characterized by a circularssDNA genomewithin twin sphericalparticles. Molecular characterization of BCTV in sugarbeethas demonstratedthat the virus primarily exists as three shains (CFH, worland, and califomia/Logan), and variants of these strains (Stenger and McMahon 1997). BCTV infectsa broadrange ofhosts that, in additionto chile,lncludesnumerous uops andweeds in manyplant families(Bennett l97l). The virus is transmittedin a circulative martnerby the beet leafiropper,Circulifer tenellus (Baker), a member of the subfamily Deltocephalinaeof the homopteranfarnily Cicadellidae.The beetleaftropper is endemicthroughout the westemand souihwestern Uj, preferring arid and semi-arid conditions. The leafhoppervector feeds and breedson an extensiverange of plant families (Cook 1967). C. tenillus exists as three morphological tlpes: a surnmer morph, a winter morph, and a migratory morph (severin tfit;. rne summer morphs survive 3-4 months, while the winter morphs live longer and consist

177 primarily of matedoverwintering females. Migratory morphsof the lealhopperare capable of flying severalhundred miles (Dorstand Davis 1937). The flight pattemsof the leafhopperin Califomia are well described,where C. tenellus overwintersin the foothills on weeds,with the nymphspresumably acquiring virus from thesehosts (cook 1967). As theseweeds dry, the mature lealhoppersmigrate into the interior valleys in the spring, feeding on and infecting crops and weeds. Spring movements of the leaflroppersin California have also been correlatedwith temperatureaccumulation (Cook 1945). The leafhoppersprogress through several generationsbefore moving back into the foothills in the fall. The current migratory pattemsof the beet leaftropperin New Mexico are not as well studied. The Rio GrandeValley was reportedto be a spring and summerbreeding area for the beet leaftropperin 1939,where they undergothree to five breedingcycles @omney 1939). The leaftropperswere thought to spreadin the spring from this areato areasnorth and east,including northem New Mexico, west Texas,and portions of Oklahoma,Kansas, and Colorado. However, the land use pattemsin New Mexico have changeddramatically sincethe 1930'swhen sugarbeet was a primarycrop, The weed host range of the beet leaftropperis well-documented(Bennett l97l), althoughdata on relativeBCTV incidencein weedsis limited (creamer et al. 1996). A perennial mustard was reported to be the most important overwintering host of the leafrropperin much of the breedingarea of New Mexico in the 1930s@omney 1939). The sameweed is now scarceat best. Given the changesin crops and weedsin the past 60 years andthe reoccurringlosses to chile dueto the virus,we felt an updateof the statusof thebeet lea{hopperin New Mexico was necessary. This paper reports the incidenceof beet leafrtoppersin chile fields in two years in New Mexico, one with high curly top disease pressureand one with low pressure.In addition,we report the diseaseincidence in weed hostsfor the two years.

METHODS AND MATERTALS

Incidenceof leaftroppersin chile was assayedby trapping insectsfrom the margins of chile fields from February2001 throughDecember 2002. h 2001, nine fields were sampled,five in Luna Countyand four in Dofia Ana County,New Mexico, while in 2002, ten fields (five from eachcounty) were sampled. Sincechile is not usuallygrown in the samefield in consecutiveyears, chile fieldslocated near those sampled in 2001were chosen for testingin2Q02. Fouryellow stickytraps (20 x25 cm) wereplaced approximately 6l cm from the ground at the marginsof eachtest field. Trapswere changedevery two weeksand leaftroppersidentified and counted. Meteorological data were gathered from weather stationslocated near the two groupsofchile fields. Degreedays were calculatedusing the singlesine method using a 50'F (10"C)lower threshold. lncidenceof BCTV in chile fields was estimatedby counting symptomaticplants in 100 randomly chosenplants in test fields and selectedadditional fields. This method underestimatesthe total amountof virus infectionin a field in a season,since symptomatic plants areremoved during crop thiruring. At two-weekintenals, weedswere collectedfrom the margin of eachchile field. After the plants were identified to species,0.5-g leaf sampleswere ground in liquid nitrogen,and total DNA extracted(Palmer et al. 1998). BCTV was amplifiedby PCRusing a viral specific primer set 5'-GTGGATCAATTTCCAGACAATTATC-3' and 5'- CCCATAAGAGCCATATCAAACTTC-3',which amplifiesa portion of the coat protein gene. Primers that specifically amplify the Worland strain (5'- CCAGGACTTAACCGCTTCATTT-3'and 5'-GGAGGCCAGCAGACGGCTAA-3')and CFH strain (5'-TCTACGTCATCAATGACGTT-3' and 5'- AGCTCCTCGCTATAAATACA-3') were used to deterrnfnethe strain of BCTV after

178 plants had been determinedto contain the virus. PCR reactionswere carried out in a total volume of 50 pl containing20.5 stlHzo, 5 pl lOX buffer (100 mM Tris-HCl, pH g.3 500 mM KCl, O.01%gelatin), I pl l0 mM dNTps, 3 1rl25 mM MgCl2, 5 pl eachof 5mM primers,.4U Taq DIA polymerase,and 10 pl of I :10dilution of purineaOf.fn. Amplification using primers to detectall BCTV strainJand BCTV-worland was carriedout with the following parameters:35 cyclesconsisting of 94oCfor 30 sec,59oC for 60 sec,and 72oc for 90 sec. A final extensionof 72"c for 5 min followed. Amplification to detect BCTV-CFH was done similarly except that the annealingst€p was done at 53oC for,60 sec' Amplification Pioductswere separatedby electrophoreiis in a 2o/oagarose gel, andthe bands visualized with ethidiumbromide,

RESL]LTSAND DISCUSSION

The beet leaftroppervector ofBCTV was trappedfrom all fields assessedduring the nearly two-year period. c. tenellusappeared in chil6 fietds in Dofia Ana county during the en! gjtvtarch and beginning of April in 2001 and2002,and in Luna county dwing early to middle May in 2001 and2002(Figs. l, 2). Leafhoppernumbers in2002pl*ra in Junein Doffa Ana county, while in runa county, the leaihoppernumbers peateo in August. tn both areassurveyed, the nym!er_9!leaftropperstrapped per field decreasedbelow 2b by the end of october 2002, while in 2001, leaftbppernrirnuerr did not drop below 20 per field until mid-December, In both years,more leafiropperswere trappedfrom Dofla Ana county fields than from Luna Counw.

- Dona Anr Co. '----- LunaCo.

gEFFEESggSFsF

D!te

FIG. l. Mean numbersof adult circurifer teneilus caught on yellow sticky traps from margin of chile tho fields in two New Meiico countiesinloot. rvot" **inio- axis. Asterisk value on i denotesdate when no data was availablefor Luna c;".tfi;;" storms. severedust

179 gREgSEF

FIG. 2. Mean numbersof adult Circatifer tenelluscaught on yellow sticky traps from the marginof chile fieldsin two New Mexico countiesin2002. Asterisksdenote dates when no datawas availablefor Luna County due to severedust storms.

The maximumnumber of leaftiopperstrapped in a field in 2001(3,500/field, Fig. 3) was substantiallymore than in 2OO2(145 leafhoppers/field).In 2001,one field locatedin Rincon, NM, happedmuch higher leaftopper numbersthan any other field from mid-April through July (Fig. 3), although diseaseincidence was not different from the other fields in Dofla Ana Countytested. Although increasednumbers of beet leaftroppersappeared in Dofia Ana County chile fields at least a month earlier than Luna County fields, the 2001 and 2002 Januaryto April temperatures(accumulated degree days) were similar betweenthe two areas. Dofra Ana Countyfields accumulate 4l loF (210'C) degreedays (DD) in 1 January-3lMarch 2001 and 398.F (203oC)DD for the intervalduring 2002,compared to 388oF(196"C) DD for Luna County for the same three months in 2001 and 442"F (226"C) DD in 2002. This also suggeststhat temperaturewas not the primary factor in determining the flight of the beet leaftropperin this study. This differs from Cook's (1945) results which suggestthat beet leafhoppersleave their breedinggrounds as soonas they achievematurity. bulng the January to April period for both years Luna county received more precipitation(1.36 itr., I January-3lMarch 2001; 1.35in., 1 January-3l March 2002)than Dofla Ana County during the sameinterval (0.99 in', 2001; 0.93 in., 2002). The extra precipitation could have delayedthe leafhoppersdeparture from overwintering weed hosts in Luna County comparedto Dofia Ana County, by delaying the drying of the weeds. Dryrng of the overwintering weed hosts is thought to be one of the primary factors that contributeto springflights (Carter1930, Severin 1933).

180 ii$t*i EE i 3iEf i I€ i dI E E

FIG, 3. Numbersof adult Circulifer tenelluscaught on yellow sticky traps from the margin of chile fields in 4 fields within Doffa Ana County in 2001. Note maximumvalue on Y axis.

The incidence of curly top in chile in 2002 (0.5-1%) was substantiallyless than in 2001 (30-50%). The BCTV incidencein weedsw.ls correspondinglylower. The incidence in weedsis likely a more accwateestimate of the actuallevel of diseasein the environment sinceinfected weeds do not showdisease symptoms, thus sampling is not biased.We found 1.3% incidenceof BCTV in weedsin 2001 (9 infected/ 686 plantstested) compared to 0.06%in 2002(l infected/1,768 plants tested). Four weedspecies were foundto be infectedwith BCTV (Table1). Thesehave all beenpreviously reported as hosts for the virus (Beruretl97l), andall havebeen reported as infected in field collections in califomia (Creamer et al. 1996). other frequently tested plants that did not test positive for BCTV infection includedbindweed (Convolulus amensis L.), ll9 plants tested;spuned anoda(Anoda cristata (L.) Schlecht),114 plants tested; ground cherry (Physaliswrightii Gray),197 plants tested;and Russianthistle (Salsolo iberica Senne& Pau),315 plantstested. The lack of infectedRussian thistle in the field, even thoughcollected for 10 monthsof the year, is similar to our findingsin Califomia where although the plant was frequently collected, it was never found to be infected with BCTV (Creameret al. 1996).

TABLE l. BeetCurly Top Virus-InfectedWeeds Collectd in 2001-2002. Species Number infected/ numbertested 2001 2002 Amaranthus sp. (pigweed) 2/237 0/174 Sisymbrium irio L. (l-ondon rocket) 2/trs 01252 Chenopodium sp. (lambsquarters) 2/57 0/125 Kochia scopaia (L.) (Schrader) 3/18 t/r95

181 All the BCTV-infected plants were found to contain the Worland strain of the virus. In addition, one chenopodium sp. was also found to be infected with the cFH strain. Infected plants were collected in June,July, August, and Septemberof 2001 and in August of 2Q02. T\e 2002virus-infected sample was collectedfrom Dofla Ana County,whili in 2001, eight infectedsamples were collectedfrom Dofla Ana county and one liom Luna County. The collectionof BCTV-infectedweeds at the marginsof chile fields throughmuch of the 2001growing season emphasizes the needfor morestringent weed control. The four speciesofinfected weedsare all reportedto be hostsfor thebeet leafiropper (Cook 1967)as well as the virus, and thus, could serveas sourceof virus for infection into chile fields.

ACKNOWLEDGMENT

we thank Marvin clary, vince Hernandez,and the New Mexico chile Task Force for their help in selectingand acquiringtest fields.We thankDayna Drollinger, Attelia Lewis, and D. P Garrido for help with extractingnucleic acid and carrying out pcR. we thank the New Mexico Agricultural ExperimentStation for funding this work.

LITERATURE CITED

Bennett,c. w. 1971, The curly top diseaseof sugarbeetand other plants. The Amer. Phytopathol.Soc. Monogr, No. 7 carter, w. 1930. Ecologicalstudies of the beetleafhopper. u.s.D.A. Tech.Bull. 206. lls pp. Cook W' C. 1945. The relation of spring movementsof the beet leafhopper(Eutettix tenellusBaker) in central Califomia to tonperature accumulations(Homoptera). Ann, Entomol.Soc. Amer. 38:.149-162. Cook, W. C. 1967. Life history,host plants,and migrationsof the beet leafhopperin the westernUnited States. U.S.D.A. Tech. Bull. 1365.l22pp. Crawford,R. F. 1927.Curlytop inNewMexico. U.S.D.A.Off. Rec.6: 8. Creamer,R., Luque-Williams,M., and Howo, M. 1996. Epidemiologyand incidenceof beetcurly top gerninivirusin naturallyinfected weed hosts. PlantDis. 80: 533-535. Dorst,H.E., andDavis, E. W. 1937.Tracing long-distance movements ofbeet leaftopperin the desert.J. Econ.Entomol. 30: 948-954. NationalAgricultural StatisticsService. 2003. United StatesDepartment of Agriculture, National Vegetable Estimation Program: http://jan.mannlib.comell.edu/reports/nassr/fruiVpvg-bban/vean0l03.txt. Palmer,K. E., Schnippenkoetter,W. H., and Rybicki, E. P. 1998. Geminivirusisolation and DNA extraction. pp. 41-52. In G. D. Fosterand S. C. Taylor (eds.)Methods in MolecularBiologn Vol.8l: Plant Virology Protocols:From Virus Isolationto TransgenicResistance. Humana Press Inc., Totowa,NJ. Romney,V, E. 1939. Breedingareas and economicdistribution of the beet leafhopperin New Mexico,southem Colorado, and westem Texas. U.S.D.A. Cir. 5I 8. 14pp. Severin,H. H. P. 1921. Summaryof the life historyof the beetleaftropper (Eutettix tenella Baker). J. Econ.Entomol. 14:433-436. Severin,H. H. P. 1933. Field observationsof the beet leafhopper,Eutettix tenellus,in' Califomia, Hilgardia7 : 281-360. Stenger,D. C., and McMahon, C. L. 1997.Genotlpic diversity of beet curly top virus populationsin thewestern United States. Phytopathology 87:737-744.

182 vol.28 NO.3 SOUTHWESTERNENTOMOLOGIST SEPT.2OO3

BIOLOGY OF THE VINE MEALYBUGI IN VINEYARDS IN THE COACIIELLA VALLEY, CALIFORNIA

K. Godfrey,J. Ball, D. Gowalez2,and E. Reeves3

California Dept. of Food and Agriculture, Biocontrol program 3288Meadowview Road, Sacramento, CA 95g32

ABSTRACT pranococcus vine mealybtg, ficus (signoret) (Homoptera: pseudococcidae), biology was studied in severalvineyards in the Ciacheila Valley'from 1994-199gand in 2000 using a variety of samplingtechniques. It was deterrnineCtirat all life stagesof VMB could be found on all pa-r.tsof the vine throughout the year, above and below ground. In spring densitiesof VMB-increaseddramaticalty, peating in mio-tate sprinf This increase i1 densitr was probably gllelo rlcreasedreproduction "id -on"*.nt orvirrre throughout the vine. Densities of vMB declined dramatically in early summer and iemaineo tow tfnoughoutthe summer,fall, and winter. Rates of farasitism on vMB were low in all of the yearsof sampling. Thosemealybugs that were exposed(e.g., leavesand clusters)were more heavily parasitized than those on more protectedparti o? the vine (e.g., roots). The primary parasites, Anagtnu pseudocolci (Girault),-were eseiaapnycus ,pp., iai"pro.astidea abnormis (Girault) (Hymenoptera. Errcyrtidae), recovered.A secondaryparasite, Chartocerusspp. (Hymenoptera:Sigrriph-oridae),' was also recovered

INTRODUCTION

In the early 1990's, grape growers in the coachella valley of califomia began experiencinga reduction in the quality and leld of their table grapesdue Ja meaybug. It was assumed that the T9?lybug causing these problemsAas the grape meaybug, Pseudococcusmaritimus (Homoptera: pseudococcidae), ^(Bhrhom) a p?.ior grapes in cenhal and northem Califomia. However, specimenssent to the California Departnent of Fgod and Agriculture-Plant pest Diagnostic Laboratory and the uS National Museum in 1994 were identified *-g: uil. mealytug, planococcasy'czs (Signoret) (Homoptera: Pseudococcidae)(cil 1994).until this identification,the vine;;l)6G ivirasj *.. hg* pest T g_gconomic of-grapesonly in the Meditenan.ro.egion of Europe,Africa and the Middle pakistan, East, South Africa, and Argentina. hr the"seareas or tne worto, vMB has also beenreportel t9 attackfig, avocado,mango, and pomegranate (Gill 1994). Management of vMB is complicatedby the cryrpticnature of-this mealybug,which spendsmuch ofits life cycle either beneaththi Uarkoithe ,rin" or on roots at the soil-air

I Homoptera: Pseudococcidae 'Dept. of Entomology, 'Retired; University of Califomia, Riverside,CA 92521 2510 Rambling Court, Riverside, CAg25O7

183 interface(Duso 1989).Therefore, biological control was the managementtactic targeted for developmentin a cooperativeproject initiated in 1994 among the California Departmentof Food and Agriculture-BiologicalControl Program,the RiversideCounty Departmentof Agriculture,and the Universityof Califomiaat Riverside.The objectivesof this project were to identify the indigenousparasitoids attacking VMB, elucidate VMB biology on grapes in the Coachella Valley, and to conduct foreign exploration and introduceexotic parasitoids of VMB. Eachagency had a specificrole assigned:Califomia Departmentof Food and Agriculture-Biological Control Program and Riverside County Departmentof Agriculture surveyedindigenous parasitoids and conductedstudies on the biology of VMB, and the University of Califomia conductedforeign explorationand introduction of exotic parasitoids.Results of the surveys for indigenousparasitoids and studieson the biology of VMB on table grapesin the CoachellaValley arereported.

MATERIALS AND METHODS

All studieswere conductedin vineyardsin the CoachellaValley from 1994through 1998 and in 2000. The vineyardssampled were selectedfor having large densitiesof vMB. The number of vineyards surveyedand the methods used to sample vMB varied with the specific objectives of each study. Although information on parasitoids was recordedover the 5-yearperiod, the survey for indigenousparasitoids was emphasizedin 1994 and 1995. From 1996-1998and in 2000, the studiesconducted ernphasized the biologyofVMB. Indigenousparasitoids of VMB were surveyedby collectingplant parts infested with vMB and holding the mealybugsin rearing containers for the emergenceof any parasitoids.From 15 November1994 through28 May 1995,collections of leaves,fruit clusters,roots, and caneswere made every other month in four vineyards.Beginning on 26 June 1995, sampleswere taken monthly until 4 October 1995,and then every other month from2/ November1995 through 19 March 1996.In eachvineyard, ten vines were selected to sampleusing systematicsampling with a random start. At eachvine, the trunk, roots, leaves,clusters (when present),and "buried canes"(i.e., canesin contactwith the soil)' when present were sampled. For the trunk, the sampling consisted of searchingfor 3 minutes by scrapingbark from the tunk and inspecting the scrapedarea for VMB. The roots of the vine were dug out, and then searchedfor 3 minutes. Sections of roots containing VMB were clipped and placed in vials. At each vine, six leaves were also collected,and ifclusters werepresent, one clusterwas collected.The soil surfacenear the sampledvine was also inspectedfor the presenceof buried canes,which were collected when found. All plant material was retumed to the laboratory and sorted.The number of VlvtB and VMB mummies (i.e., parasitizedVMB) were recordedfor eachplant part and sampledate. The VMB were placed in rearing containerswith a sproutedpotato to allow the developmentof any parasitoidsthat may have beenpresent. The VMB mummieswere placedin vials and held for emergenceofthe adult parasitoid. The biology of vMB was investigatedfrom 1996through 1998 and in 2000 by assessingVMB density and activity on various plant parts using a variety of sampling methods.In 1996,1998, and 2000,all studieswere conductedin one vineyard,a certified organicblock ofSuperior Seedlessgrapes. In 1997,studies were conducted in the certified organic vineyard and in a block of conventionallymanaged Thompson Seedlessgrapes. Foi all years,VMB mummies that were found were collected, retumed to the laboratory, and held for the €mergenceof the adult parasitoid. The number and identity of the parasitoidswere recorded. The biology of VMB in 1996was investigatedusing three setsof 12 vines that were sampledmonthly from 14 May through 19 Novemberand at bimonthly intervals

184 from 19 Novemberthrough 2l January1997. The first set of 12 vines was randomly selectedand sampledrepeate_dly. on 19 March 1996,loose bark was removedfrom the trunk of eachvine, a 250-crl areamarked on the exposedtrunk, and then coveredwith a layer of burlap. kr addition, a section of a small root (l-2cm diameter) was exposed, loosely wrapped in a sheetof plastic, and reburied. The intention of thesemanipulations was to delimit specific areason the vine for repeatedsampling begrnning 14 May, while providing physical protection from predatorsof vMB. on eachsample date, the coverings on the trunk and roots were removed,the number of VMB and mummiescounted in situ. and the coveringsreplaced. In addition,eight leaveswere collectedfrom eachvine and retumed to the laboratory. The number of vMB and mummies found on each leaf were recorded, The secondset of 12 vineswas immediatelyadjacent to the first set andwas used to investigatethe effectsof artificially coveringthe exposedtrunk. For thesevines, the bark was removed from the hunk; a 250-cmz area was marked on the hunk, but no covering was placedover the exposedarea. on eachsanrple date, the numberof vMB and mummieswere countedin the delimited areaon the hunk. The third set of 12 vines was used in an attfiryt to track VMB movement tlrroughout the vine. On each sample date, 12 vines were randomly selectedand the bark removedfrom three areasofthe cordon and greericane (i.e., current year's growth). These areasiwere proximal, medial, and distal to the trunk. In theseareas, sticky traps consisting of two widtbs of double-sidedsticky tape(width of tap*l.9cm) wereplaced encircling a cordon or cane.on eachsample date, the traps were rernovedfrom the field. The number of eachsize classof vMB on eachtape, and the length of eachtape were recorded. Studieson the biology of VMB in 1997were conducted on l8 vinesin the certified organic vineyard and six vines in the conventionally managedvineyard beginning 24 February 1997 with sampling at monthly intervals from 2 April through 24 Novernber 1997. The vines were specifically selectedfor having observableVMB infestations in January.on each sampling date, one side of the tnrnk (from the soil line to about 30cm abovethe soil line) was examined,and the number of VMB and mummiesfound recorded. In addition, ten leaves,one bud, and one cluster, when present,were collected for each vine and returned to the laboratory.The numbers of vMB and mummies found on each plant part were recorded.In the organicvineyard, the roots on six vines, selectedat random from the 18 sample vines, were exposed,and the number of vMB and mummies found recorded.No roots were examinedin the conventionallymanaged vineyard. To investigatevMB movement,sticky traps were placed aroundan arm, a cordon, a brown cane (i.e., previous year's growth), and a green cane on each vine. Each trap consistedof a singlewidth (l.9cm) of double-sidedtape. The tapeswere replaced on euch sampledate. The number of each life stageof VMB and mummies on each tape and the length ofthe tape were recorded. In 1998, all studies were conductedon 12 vines that were selectedfor having a readily observablemealybug infestation in March. Two types of sticky traps and refuge areaswere placed on the vines on22 Apil and monitoredthrough 26 August 1999.on eachvine, a singlewidth (1.9cm)of double-sidedtape was placed around an arm, cordon, and brown cane.In addition,a yellow sticky trap (7.6cmx l2.7cm) was hung underthe canopyofeach vine. The sticky tapesand yellow trapswere replaced evt.lry 2 weeks.The number of each life stageof vMB and mummies on eachtape, and the length of the tape were recorded'The numberof male VMB and the numberand speciesof parasitoidsfound on eachyellow trap were recorded. Refugeareas were also establishedon rm arm, cordon,and brown caneon eachvine opgosite the sticky tapes.Each refuge consistedof a 5.08-cm wide strip of bubble wrap (-3.5 bubblesper cm2)wrapped around the appropriateplant part and lield in place witir

185 duct tape. When initially establishingthe refuge areas,a seriesof three wraps was placed on eachplant part on eachvine. Each wrap in a serieswas assignedat randomto a 6-week, l2-week, or l8-week samplinginterval. Thosewraps assignedto the 6-week interval were replacedwith new wraps when removed for the next 6-week samplinginterval. The wraps that were removed were placed in a vial along with any mealytugs adheringto the vine undemeaththe wrap and returned to the laboratory.The numbersof VMB and mummies found under eachwrap were recorded. In 2000,VMB biolory was investigatedusing sampling methods that had beenthe most successfulin previousyears. This included field counts,sticky tape traps,yellow sticky traps, and refugeareas. On 24 February,270vines with active VMB infestations were selected and randomly assignedto one of three groups: field counts (50 vines), trapping (20 vines), and refuge areas(200 vines). Samplingwas conductedmonthly from 24 Februarythrough 7 November2000. The field count methodwas usedto investigatechanges in the VMB population on different parts of the vine through time. At the beginningof the study,each vine within the 50-vine group was assignedat random to one often subgroupsto be usedon one sampling date. All vines in one subgroupwere examinedfor VMB on the trunk, cordon, and roots. The VMB and mummieswere collected, retumed to the laboratoryfor counting, and held for parasitoid emergenceand identification. The number and size class of VMB, the numberof mummies,and the parasitoidspecies found on the variouspads of the vine were recorded. The trapping method was used to investigateVMB movementthrough spaceand time, and to determinethe general activity of VMB males and parasitoids.The same20 vines were sampledthroughout the study. In February,the loose bark was removed from portions ofthe trunk and one cordon on each vine. A sticky tape hap was applied to the exposedareas. A yellow sticky trap was hung in the canopy ofeach vine. All traps w€re replaced monthly. The number of each size class and/or life stagesof VMB and the numberof eachVMB parasitoidspecies were recordedfor eachtrap. The refuge areasiwere used to investigateparasitism of VMB. The 200 vines were assigrredrandomly to one of ten subgroups.Each subgroup was used on a different sampling date. On each sampling date, the loose bark was removed from an area of the trunk andcordon on eachvine, and a refugearea (i.e., band ofbubble wrap)was appliedto the exposed area. After one month, the refuge areas were removed, retumed to the laboratory, and the size class of VMB, their numbers, and the number of mummies recorded. The VMB and mummies were then held for parasitoid emergenceand identification, The countsof VMB on various plant parts were summarizedas meansand standard errors per plant part for each sample date and vineyard. The sticky trap data was summarizedby calculatingthe meansand standarderrors for VMB per cm'of tapeon each plant part for eachsample date and vineyard. The yellow sticky trap and refuge data were summarizedby calculatingthe total number of VMB, mummies,and parasitoidsrecovered for each sampledate. The levels of parasitism for VMB from the field counts and refuge areaswere calculatedas the ratio of the number of mummies (intact and emerged)to the total number of VMB collected (expressedas percentage)for each sampletype, sample date,and vineyard.

RESULTSAND DISCUSSION

The preliminary survey of indigenousparasitoids attacking VMB in the Coachella Valley revealed low levels of parasitism (Table l). The primary parasitoid Anagnts pseudococci(Girault) (Hymenoptera:Encyrtidae) was recoveredon all sample dates for

186 which parasitismwas detected.From the samplescollected 6 Septemberand 27 Novernber (Hymenoptera: 1995, another parasitoid belonging to the genuri Chartocerus_ Sig;phoridae) an-dthought to be hlper-parasiticwas recovered.On 6 September1995, liyo otmeparasitoidsrecovered were hlperparasitoids,'and 14.60/owerc hlperparasitoids on 27 November1995. Althougb generally low, the amount of parasitismvaried with the exposureof the mealybug.For-VtvtB with few placesto hide such as on leavesor clusters,parasitism rates were as ttigtt ^ 22Yo and l3olo,respectively (Table l). For thoseVMB that were hidde r or protected,such as on roots or bwied canes,parasitism rates rarely exceeded1% (Table 1)'

TABLE l. Parasitismand Total Numbers of VMB on Different Parts of the Vine in the - CoachellaValley in 1994 1996. , , PercentParasitis@ Date l5 Nov I 0.1(7ll) 0.4 (6,983) 0.7(143) 16Jan 1995 O(276) NS NS NS 28 Mar 1995 0 (1,379) NS 0 (16) 0 (104) 23May 1995 0 (314) NS o(r4) 0.3(17e) 26 Jun 1995 0 (36) 0.4(510) rz.e(74) 2.4 (40) 2 Aug 1995 0 (117) 0.2(1,367) e.8 (37) 22.0(8) 6 Sept1995 1.2(81) 0.4(4,s48) NS 0 (15) 4 Oct 1995 0 (59) 0.3(3,060) NS 5.3(18) 27 Nov 1995 0 (458) o.t (1,224) NS 0 (18) 23 Jmr996 0 (14) NS NS NS 19Mar 1996 0 (ll2) NS NS 0 (4) f.iS : No sampletaken

The pattems of VMB seasonalabundance were similar for all sampling methods over the four years of study. Field counts of VMB on various parts of the vine in the organic vineyard in 1996, 1997, and 2000 began to increasein late May through June, p"-rking in May or June (Figs. l, 2a, 3). Densities of VMB generallydeclined during the .o--Jt with a slight increasein densityin Septemberand October(Figs' l' 2a). Therewas a secondpeak in densityin August in 2000 (Fig. 3). For this peak,three out ofthe five vines sampledhad very large densitiesof VMB, and two had no VMB. This demonstrates the chmfed nature of VMB distribution within the vineyard, and, consequantly,the limitations of data from only a few samples. The counts of VMB on various plant parts also revealedthat there was a resident year -(FUr.population ofVMB on the tunk, and to a lesserdegree, on the roots throughoutthe 1, 2a, 3). The extent of VMB habitation of roots is influencedby many factors such as ant activity, soil texture, temperature,and irrigation regimes. Even in Januarywhen vines were pruned,VMB could be found throughoutthe vine (e.g.,trunk, roots, the baseof buds, pruning scars, etc.). In addition, reproducing VMB females could be found throughoutthe year (J. Ball, unpublisheddata). As the vines developedcanes, leaves, and clusters in the spring VMB moved outward from the main stem or tunk colonizing the newplant parts (Figs. l, 2a, 3).

187 ...*-Wrappod Trunk

- l- ExposedTrunk

o -{-RootB E o -t(-Leaves o z o o

r4.tfiy l&Jun t}Jut 12.A|.A tl\Sep 1t.Oct lo.Nov ,GDec *Jan Date

FIG. l. The meannumber of VMB found otr the various parts of the vine in field countsin I 996. The barsrepresent the standarderrors of the means.

The trend in VMB density in 1997 on vines in the conventionally managed vineyardwas similar to that seenin the organicvineyard in (Figs. l-3). The densityof overwintering VMB in Februarywas much lower in the conventionallymanaged vineyard (Fig. 2b). In addition, this vineyard was treated with methomyl in late March, further reducing vMB densities in April (Fig. 2b), From May through August, vMB densities resurged,followed by a declinein Septemberthrough Novembir (Fig. 2b). No vMB were found on leavesor clusters,possibly due to the lower density of Vivf6 on the trunk early in the season. In the field counts in 2000, parasite mummies were recovered from February through october. No mummies were recoveredfrom the root samples. The parasitel. pseudococciwas recovereda few months prior to the recoveryof Leptomastidei abnormis (Giraul0' Hymenoptera:Encyrtidael. Both parasiteshad beenreleasld within the vineyard from 1998through 2000, but outsidethe studyplots. The collectionsmade on sticky tapetraps in 1996-1998and in 2000demonstrated the samepopulation trends as those found using the direct count method.In general,there was an increasein the number of vMB happed throughout the spring with a decline occurring in mid-summer.There appearedto be a residentpopulationbf VMB on the trunk t99ugh.{ most of the year with movementoutward to other plant parts during the spring (Figs. H). In the conventionallymanaged vineyard in 1997,the movementoutward was delayedslightly due to the applicationof methomyl in late March (Fig. 5b). The trap catcheson the yellow sticky trapsplaced in the vineyardin 1998and 2000 show pattemsconsistent with the field count and sticky tape hap data. vMB male flight peakedin mid-Julyin 1998,and in April in 2000 (Fig. 7). The peakin vMB male flights occruredabout I month after the peak trap catch for the younger life stages(Fig.6). The delay in developmentof vMB males in 1998compared to 2000 was due io a much colder than normal spring in 1998, and a warmer than normal spring in 2000 (www.ncdc.noaa.sov).

188 . l- Trunk *Rools -l(-Leawe € zo.oo .-+Bttd "+...t'..-a- .clust€r o , 15.00 zd g 6 $ ro.oo t

I I .'{ \ "'lt' ,./ Et , a

2&F€b 2&Uar 25-Arr 2erfty a^!rn aLJU aArrg 22's,€p AZ'oA zl'rbt Date

- l- Trunk ao 20.00 = ..,#Bud o zo r!,00 c o =o "'-{- a ,'1.... t

.,,{-- o.m 2a-Fab ?!|'-f,/g,t b pt b)Et 21-Jwt 24.16 ?3',,4g Z}$q Zz'()d 21't{@'t Data

FIG. 2. (a) The meannumber of VMB foundon variousparts of the vine in field countsin an organic vineyard tn 1997. (b) The mean numberof VMB found on various parts of the vine in field counts in a conventionallymanaged vineyard in 1997.Methomyl was applied U in March. For both gaphs, the bars representthe standarderrors of the n#:.vr""t

189 The yellow stickv haps also trapped the adults of the two parasitesthat YMB' Anagtruspseudoioc^ci'was attack found'ontrre traps in both yearsbefore L. abnormis. 199€,peak trap catches In for both parasitesoccr.ileo in July, whereas,-'-'--e, A."' pseudococci peakedin May andZ. abnormis I peakid in Junein 2000Gig. Zj.'

= > 60.(X) o 50.00 zd E 40.00 0) E go.oo

20,00

t0.00

0.00 n-Jvt 2rJ!l

FIG' 3' The mean number of VMB found on various parts of the vine in field counts in 2000. The barsrepresent the standarderrors ofthe means.

ge are:rsplaced ,- . . rh:r"n on the vinesin l99g and2000 did harborparasitized vMB (Table2)' Refugeareas left in the field over shortperiods of time allowedbetter detection of the parasitesbecause parasitized the vMB and/or the evidenceoip**iti* was not with the p.rssage :t:I?y"d of time (Table 2). The parasitesfrom the retugeareas were rouno at approxrmately the sametime as the adult parasiteswere found Jn the yellow stickycards (Table 2 andFig. 7). over all years of sampling, the rates of parasitism on vMB were reratively low. Those vMB that were more exposed (e.g., on leaves and clusters)'*J were more heavily parasitized than those in proiected loiations (e.g., on roos t*ied canes). Ay-qnerltatiol of the parasite populations did increasethe amount orprr itir* on vMB within the releasevineyard. combining the data from all years and all sampling-datiwere techniques, the biorogy and seasonalphenology of VMB were pieced together.The consisientwitfr the idea tl".e was a {ut residentpopulation of vMB on all parts of the vine throughoutthe year. Even in winter when vMB was at its lowest density, it could b, fo;d hiil on the vine tu.ft under T. bud.scales(J. Ball, unpublishedcata;, ana some orme aeriaffiuhtion probably was removed during winter pruning. firere was no winter aestivation or diapause becauseall life stagesof vMB could be found throughoutthe year. In fi;;, densitiesof

190 0.800 E 0.700 o C' U' 0.600 0) o 0.50{) (D o.400 o 0.300 zct c 0.2{x, o o = 0.r00

0.000 l&Jun l6Jr,l 12.Ano Date of Trap Placsmenl

4.000

E 3.500 C) ,f r.ooo E r.u* co I z.ooo 9 r.soo o z 't.000 c o o > 0.500

0.000

Datoof TrapPlacement

FIG. 4. (a) The mean number of VMB per sq cm of sticky tape trap placed on canesin a vineyardin 1996.(b) The meannumber of VMB per sq cm of sticky tapehap placedon cordonsin a vineyard in 1996.For both graphs,the bars r€presentthe standarde,lror ofthe means.

191 14.C0

E r2,oo o (a io.oo ACordon o 8.O0 =6 g2rd yrcane 6.00 ts B lsl yr C6ne ci z 4.0t)

$ z.oo

IJln 15{ul 1SA!g Date of Trao Placsrnonl

E (J 3.00 g a, k 2.50 $ a.oo

o 1-50 z.^o 0 E o.go

?4-FeD ?-Nt 5.May g\,l]n l$Jul 19.4il0 1$Ssp 27-Oct 2.4-ttov Dale of Trao Placemenl

FIG. 5. (a) The mean number of VMB per sq cm of sticky tape trap placed on various parts of the vine in an organic vineyard in 1997. (b) The mean number of VMB per sq cm of sticky tape kap placed on various parts ofthe vine in a conventionally managed vineyard in 1997 . For both graphs, the bars represent the standard errors of the means.

r92 l

6.00 E o 5.00 E o to u't

b 3.oo zo c 2,N ct o E 1.00

zilay 3^Jun l&Jtn &Jul ltJul 3t.Jul 14-Aug 26.A',/€ Date of Trap Placement

5.00 4.50 lTrunk 6 4.m (t a' 3.5o o ZCordon I r.oo 3 2.so ? z.oo 7'r.5o tg E Loo 0.50 0.00 24-Feb 27-Mar 24-Ap( 23-May 21Jun 24-J!l 3l-Aug 2'od Dale of Trap Placement

FIG. 6. (a) The mean number of VMB per sq cm of sticky tape trap placed on various parts of vines in 1998. (b) The mean number of VMB per sq cm of sticky tape trap placed on various parts of the vine in 2000. For both graphs, the bars represent the standard errors of the means.

193 80.00 70.00 E, F h 60.00 o- g 30,00 3' o g 4o.oo ? so.oo zo cZoffi o o = 10.00

0.00 Z-|tlay .l-.run t8\run }Jd 1&Jul 3hXrt Dateot TrapBacernent

F. 25.00 F E ,0.* o E e ts.oo o f ro.m c o € too

24-A{r 23-May Z1"ruo 244d 31"A4 Date of Trap Placenrent

FIG. 7. (a) The mean number of adult insectsper yellow sticky card placed in the vine canopyin 1998.(b) The meannumber of adult insectsper yellow stickycardplaced in the vine canopy in 2000. For both graphs,the bars representthe standarderrors ofthe means.

194 l

TABLE 2. Mean Numbers of VMB and Total Numbers of ParasitesRecovered from RefugeAreas in 1998and 2000. Date MeanNo. of VMB per sq cm TotalNo. of Refuge (r s.E.) ParasitesRecoveredo in Field Trunk Cordon Cordon Cane

1998 Apr - Jun 0.l5 0.14 0.67 lAnag& 0 0 (r 0.05) (r 0.06) (* 0.28) 3 Pseud Jrm - Jul 1.30 0.35 0.29 3chaxt 2Anag& 16Anag & (r 0.35) (r 0.10) (t 0.08) 6 Chart 4 Chart Jul - Aug 0.002 0 0.003 3 Chart 0 0 (i 0.002) (r 0.003) Apr - Jul 0.004 0.003 O.O2 l Anag 0 5 Anag (+ 0.002) (r 0.002) (r 0.007) Apr- Aug 0 0 0000 2000 Feb - Mar 0.01I 0.008 (r 0.007) (10.007) Mar - Apr o.o2 0.04 0 I Anag (r 0.006) (r 0.01) Apr - May 0.04 0.02 I Lept& 1 Anag (+0.01) (r 0.006) 4 Anag May- Jun 0.006 0.012 2Anag& lAnag (r 0.002) (10.008) I Chart I I-ept& 3 Chart Jun - Jul 0.002 0.005 00 (r 0.001) (r 0.0005) Jul - Aug 0.02 0.006 0 0 (r 0.02) (t 0.003) Aug - Oct 0.0004 0.0004 0 0 (* 0.0004) (r 0.002) Oct - Nov 0.0004 0.003 0 0.0004 Anag = Anagtrus pseudococci adults; Pseud = Pseudaphycusspp. adults; Lept Leptomastideaabnormis adults; Chart = Chartocerusspp. adults

VMB increaseddrarnatically, peaking in mid-late spring. This dramaticincrease in density was probably due to increasedreproduction and movementof vMB throughoutthe vine. In early to mid-summer, the densities of vMB declined dramatically and remainedlow throughoutthe fall and winter, The role of the roots on VMB seasonaldynamics requires further study becausemany factors may impact that role. A similar paftern in phenology was reportedby Berlinger (1977) for VMB populationsin southemIsrael.

195 ACKNOWLEDGMENT

The authorsacknowledge M. waggonerand S. o'campo for assistancein the field, S. Triapitsynfor identificationof the parasites,and R. Gill andK. Daanefor reviewingan portions gli:r at"! of this publication. of this study were funded by a grant from the CalifomiaTable Grape Commission.

LITERATURE CITED

Berlinger, M. 1977.The Mediterraneanvine mealybugand its natural enemiesin southem Israel.Phyoparasitica 5: 3-14. planococcus Duso, c, 1989. Bioecologicalstudy on ficus (Sign.) in veneto. Filippo Silvestri.46:3{:0. Gill, R. 1994.Vine mealybug.California plant pestDis. Rpt. 13: g-9.

196 vol.28 NO.3 SOUTHWESTERNENTOMOLOGIST SEPT.2OO3

LYGUS' SPECIESASSOCIATED WITH TEXAS HIGH PLAINS COTTON. ALFALFA AND WEEDS

J. ScottArmstrong and L. De AzevedoCamelo

Departmentof Plantand Soil Science,Texas Tech University and rexas Agricultural ExperimentStation, Lubbock, TX

ABSTRACT

A preliminarycollection of lygus speciesfrom cofton, Gossypium hirsutumL., onthe TexasHigh Plainsin I 999indicatedthat Lygus hesperusKnightand Lygus elisus van Duzee werethe most common, economically damaging species, followed by very low densitiesof Lygus lineolarls (Palisot de Beauvois) andPolymerus basalu (Reuter).Further intensive surveysin 2000and 2001revealed that duringJune and July the probabilityoffinding l. hesperusand L e/isnsin productioncotton or alfalfaMedicago sativaL. was essentially the same.Alternate weed hosts such as kochia, Kochia scopana L., lambsquater,Chenopodium album (L.) Schrad.,yellow sweet clover,Melilotus fficinalis, L. and redrootpigweed, Amaranthusretroflexus L, wereacceptable hosts for both Z. hesperusand L. elisusand also Polwerus basalis(Reuter), a relatedspecies that feeds mostly on weedflorets dnd is not a damagingpest of cotton.

INTRODUCTION

Lygusis the genusname for a groupofclosely relatedplant feedinginsects in the family Miridaethat have a broadhost range of weeds,legumes and cultivated crops such as cotton,Gossypiurn hirsutum L, This groupof insectscan be devastatingto cottonyield becausethey feed specifically on thenewly developing squares, anthers and small bolls. The cottonplant can compensatea portion of yield potentialuntil it becomestoo late in the seasonfor squaresand bolls to developand mature.Cotton is most susceptibleto Lygus feedingfrom the time it beginsto developsquares to peak-bloom,although young bolls are susceptiblelate in the season.There are over thirty Nearcticspecies of Lygus in North America(Schwartz and Foottit 1998)but not all aredamaging to cotton.Identiffing rygru to speciesis very difficult, andfew entomologistcan make positive identifications without the aid of a mirid taxonomist. Within the cottongrowing regions of the UnitedStates, the mostuniversal species found in the southeastemregi onsis Lyguslineolaris (PalisotdeBeauvois) also known asthe tarnishedplant bug; whereas, in thewestem cotton production regions the western tamished plant bug, Lygus hesperusIftright, is cited as the most economicallythreatening species to field cropsand specifically cotton (Layton 2000). Less frequently cited mirids ofeconomic concemin thewestem production regions are the pale legumebug, Lygus ellsus van Duzee

I Hemiotera:Miridae r97 (McGregor1927, Fye 1975,Sevacherian and Stem 1972, Kameret al. 2000) andLygus desertusKnight (Stitt 1949,Clancy 1968). Lygushave become more of a significantpest of cottonon the TexasHigh plains overthe last four yearsalong with L. lineolarisin thesoutheastem cotton production regions (Layton2000). The TexasHigh Plainsproduces between 1.2 to 1.6million ha of cottona year,which accountsfor about60% ofthe totalproduction for thestate (Sansone et al.2002), in a shortproduction season. Many entomologistsbelieve that plant bugs and stink bugshave becomemore important insectpests because of recentchanges in cofton pestmanagement strategies.These changes include the utilization ofBt (Bacillus thurgien^r,s) cotton varieties, theavailability oftarget specifrc insecticides, and boll weevileradication. Fewer insecticides areapplied in earlyseason, which allows for plantbug reproduction or continuedmovement into cottonand ultimately results in moredamage. The mirid complexthat is damagingto TexasHigh Plainscotton has not been identified. The purpose ofthis researchwas to identifi the rygzs complexin cotton,alfalfa andassociated weed hosts dwing the springand summer of 1999through 2001. The result ofthis surveyshould help identify lygzs speciesthat damagecotton and differentiate from those speciesthat may be related but do not cause damage.Some mirid speciesmay be incidentalin cottonbut their identificationwill resultin lessconfusion and possibly fewer insecticides applications.

METHODSAND MATERIALS

Lygusspp. were sampled from twelveproduction cotton fields that ranged from23-60 ha in I 999from Lubbock,Lamb, Castro and Swisher counties to determineif speciesother thanL. hespenrsinfested Texas High Plainscotton (Fig l). Fifty sweep-netsamples with a standard38-cm diameter sweep-net were taken along with samplingl5-m of linearrow with a KISS sampler(Beerwinkle 1998) within eachfield. From the 1999collections, it was determinedthat three speciesfrom the genusZygr.rs and one relatedspecies from the genus Polymeruswere found in cottonand warranted a broaderhost survey of othercrop species and weed hosts..Lygas were sampledfrom all countiesin the TexasHigh Plainswhere cotton,alfalfa, peanuts, and sunflowers could be foundduring the months of May to August for 2000and 2001 seasons (Fig. 1). Weedhosts were also sampled along crop borders and roadsideswhen they were present around cultivated species. Sampling was conducted with either a 38-cm sweepnet for the broadleafweeds or KISS samplerfor crop species.A minimumof 50 sweepsamples and a minimumof l5-m of linearrow weresampled from crops.The contentsof the samplerbag were transferredto plasticself-sealing bags and placedin afreezeruntil the adultscould be pinnedand identified. Nymphs ofall species werecollected and counted for all samplesbut not identifiedto species.The proportions of eachspecies from the total collectedon a givenhost plant were calculated. For cottonand alfalfa,the surveysamples were consistent enough from eachcounty that the distributions wereplotted for themonths of May, June,July andAugust. lygns distributionsfrom weed specieswere not plottedover this four monthperiod because weeds were a viablehost for only a few weeks.

RESULTSAND DISCUSSION

Four mirid specieswere collected in cottonin 1999:Lygushesperus, L. elisus, L. lineolaris, andPolymerus basalis (Table 1). The proportionsofadults was in favor ofL hesperus,followed by L. elisusand lessthan five specimenseach ofl. lineolaris andP. basalis.Lygus hesperushas beenconsidered the main Lygus pestof TexasHigh Plains

198 cotton,however, little is know or mentionedabout other species such as L. elisus.Lygus lineolarisis consideredan eastempest ofcotton but doesoscur in thewest (Anderson and Schuster1 983, Schwartzand Foottit 1998).Polymerus basalis has been recovered from the Mississipi Delta from weedhosts of the family Asteraceaebut is not considereda cottoirpest (Snodgrasset al.I 984). Alfalfa and canolawere dominatehosts for L. hesperusin 2000 regardlessof the samplemetlod used,followed by only afew L. elisusand L. Lineolaris (Tablel.) Cotton sampling resulted in a high number of individuals with a greaterproportion of L. elisus (0.58) comparedto Z. hesperus(0.40) during 2000 using the KISS samplerbut sweep sampleswere all hesperus(Tablel).

Fig. I . Countiessampled for mirid specieson theTexas High Plains. A's representcounties sampledfor cottonin 1999,B's werecounties where cotton, alfalfa and weed hosts were sampledin 2000,and c's arecounties sampled for cottonalfalfa and weed hosts in 2001.

com or peanutdo not appearto be an importanthost ofZygns species,although corn has beenimplicated as an importanthost for L. lineolaris southeasi(young Daq, ana peanuthas been suspected as being a goodhost from thesouthern high plainswhere peanut acreagehas expandedsignificantly over the last ten years.

r99 oo\ Zo\ ql c.t al ^-A^ ot 5 iia)i){)<.x=OOoo ot * AA -iAA-i,-i-{^' o^ (.)l vvvveeeCjcjCj 9E (t)l 7)o0 I -c{ AIAAA t- x=><)<>-lxx vlF< -iYYY'9e9OO- .:. >, AA :^^ H(V vue--- 3>

:d 7A Or €'g Hq,tr s_= 988o.38oSo0 E* iJ!;\ F.- R IEn* Hs* KFE EE Frv o\ 3# O\X {'o N dE u3 ooo c.\lhXr|OhXrlr'66.=3-.-OiaO 466 :^ 9\. l.'.X - oo 6:.,: F- J = dl htr --:-> 3E NN \O-+-\OclahciC{!f, za N..F \o (DE JO 9() ()'l' p. g.. - ^ 6 A A A 3A 8 F NggAFAFT.^ e UEE .3E =0) s (/) d> ir; zzEzEitEEAAA %A F^O 1.9) j A}

qX , fr*? NH sE T! j gHx=

;zFV o Q9 cD .a ca EE rr.. sst*?€Fi 6 a

I^ G It g$$$,E*$siiiq Xo qN€33gF*tg**si Eg g.Eu** gS ta 6 S€ 6 RdrBi,?i'i tr;f Aqr rrl O Egs 770)ess33issF$re5$g v2c-.1 oo

200 -9NQ O QQOo a A O OOoOO oqqio o ,999Q , q a E O-5-6 oooo o oooo ------

OOOr

\OOOco!>O€caO\O\O iOoOoosl 1QQN v') 'Qang , v.tq i \OC-C!h\O ocjcjd o oood o o o A -i -i -i ^i

r.r o co oo OOOOh E,BqEE ,99393s$g cjJcjo d -:ocj-j o o 6 dd3;J

a ^A c! F-= 6 ^ e SooS -roee._^ o = t6*f\- : $oo f i *=3== i\ \af S 6 c = ti-

N^ 9'AA9 - A A a A a AA' 355KaE'i Saga-88oeb' N -: o\ h - € - I g 3 9e3.33 :yv-: \o R =:g-Y c{ oo..r9Io.+n

FSr E}J}S gg;ggggg# gggg;g av2-rt)

j :H: ll ir-{*i A* ? ;: iFFF -ES S€ 33sy aF i i $;-{s 3EE F rI]$cuc"r$NS Z:t 8€ ':FEE: SqEE9 -:.8 E E I g$sgFg$,gB$giii$g He68gts-o O'

201 Weedspecies sampled in 2000favored L. hesperuswith the exceptionof redroot pigweedand a mixedsweep sample of ragweed,Russian thistle and kochia where P. basalis (Reuter)made up >80%of theadult collections (Table I ). Yellow cloverserved as the most diversehost of the weedspecies with a significantnumber of nymphscollected. The 2001 samplesfrom alfalfa were predominatedby L. elisus comparedto z. hesperus,which was a completecontrast to the 2000 sampleswhere z. hesperuswas the dominatespecies. Com again did notprove to bea hostofZygls. Theratios ofadtlt L. elisus arrdL. hesperuswere almost identical when collected from cottonin 2001(Table l). Peanut hasbeen implicated as a goodhost for Lygusand apossible reason for increasedproblems on the High Plainsin cottonbut from limited samplingfor two years;only an occasionall. hesperuswas collected. Lygus elisus was the most prevalent species on mostweeds in 2001 with the exceptionof lambsquarterand yellow clover whereL. hesperuswas in higher proportions(Table 1). No otherLygus or relatedspecies were collected from weedsin 2001. The seasonaldistribution of Lygus collected from cotton and alfalfa show some similar trends (Fig. 2). May was dominatedby higher proportions of L. hesperus;J:uuirc

Nl Lnrcp"rr*l--l e-ri,irrs J L.m"ot"rk

3 n=9 n=330 n=388 n=296 E roo e90 r7n980 : 6rl €50 s40 =JU *20 LIU 0 g n=365 n-241 n=305 n=106 g 6 100 Eso ;80 g- -^ru J60 OAA v4u F30 Ezo toFro May June July August

Fig. 2. Seasonaldishibutions ofthe percentageof Lygusspecies collected from cotton(top graph)and alfalfa (bottom graph) surveyed monthly from TexasHigh Plains counties forthe 2000and 2001 production seasons combined.

and July were usually in favor of L. elisus,and August was again dominatedby I hesperus.Alfalfa was responsiblefor generatinghigh numbersof May; whereas,few Lygusof eitherspecies were collected in cotton. Cottonis plantedfrom earlyto lateMay and is not producing fruiting forms or flowers to attractLygus. Lygus lineolais is found only occasionallyin cotton or alfalfa.

202 The 1999survey from cottonand the 2000-2001survey ofcrop speciesand weeds clearlyindicatedthat L. hesperusand L. elisuswere the damagingspecies in field cropson the TexasHigh Plains.L. lineolariswas found only occasionallyin cottonor alfalfa and usuallyin July or August,Polwerus basaliscould be foundin associationwith broadleaf weedsand was included in this surveybecause it canbe misidentified as a damagingspecies. The results of this trygus srrrveywere different from any other cotton production regionin the UnitedStates infhat L.hesperusor L. e/isascould be foundinfesting cotton or alfalfain similarproportions. Zygr.rs lineolaris is reportedas a tlueateningcotton pest in the NorthernBlackland Prairies of Texas(Womack and Schuster1987) while L. hesperusis recordedas the prevalent species threatening cotton in southwesternOklahoma (Karner et al. 2000).This is thefirst reportwhere L. hesperusand L. elisusare equal in termsofhost plants anddistribution. The proportions of thesetwo arevery similarwhen cotton was setting fruit in Juneand July. Ifinsecticides are used as a controlmeasure, it couldbe important to know which speciesis presentbecause the efficacyofa giveninsecticide may not be the samefor thetwo species.Insecticide efficacy and insecticide resistance are documentedin L. hesperus (Grafton-Cardwellet al. 1997),but no informationif anyis availablefor L. elisus.

ACKNOWLEDGMENT

We thankGye Kraemer, Li Hoabin,Scott Russell, Brant Baugh, Phillip Kidd, Clyde Crumley, Andy Cranmer, Brian Henson, Daniel Hooten for sampling insects and Joe Schafftrerand Michael Schwartzfor mirid identification. This work wassupport in partby a Texas Departmentof Agriculttre IPM grant. This manuscript is contribution T-5-530, Department of Plant and Soil Science, College of Agriculture Sciencesand Natural Resources,Texas Tech University.

LITERAruRECITED

Anderson,R.A. andM.F. Schuster.1983. Phenology of the tamishedplant bug on natural hostplants in relationto populationsin cotton.Southwest. Entomol. 8: 131-136. Beerwinkle,K. R. I 998. Fieldevaluations ofthe KISS,a tractormounted sampler and hand samplerfordetectingbollweevils inpre-bloomcotton. Southwest. Entomol.23:351- 359. Clancy,D.W. 1968.Distribution and parasitization of someLygus spp. in westemUnited Statesand central Mexico. J. Econ.Entomol. 6l:443-445. Fye, R.E. 1975.Plant host sequenceof major cottoninsects in southernArizona.USDA- ARS-W-24:9 pp. Kamer,M., J. Goodson,and D. Arnold.2000. Field surveyto assessIygas spppopulations in Oklahomaagri-eoosystems and potential pest status, p. I099-I100.ft Proceedings BeltwideCotton Conference National Cotton Council, Memphis, TN. Grafton-Cardwell,8.E.,L.D. Godfrey,W.A. Brindley, P,B. Goodell. 1997. Status of Lygus bug and cotton aphidresistance in the SanJoaquin Valley. p. 1072-1Q74.InProc. BeltwideCotton Conf., New Orleans, LA. 7-10Jan. 1997. Natl. Cotton Counc. Am., Memphis,TN. Layton,M. B. 2000.Biology and damageof the tarnishedplant bug, Lyguslineolaris, in cotton.Southwest. Entomol. 23:' 7 -20. McGregor,E.A, 1927.Lygus elisus: a pestof thecotton regions in Arizonaand Califomia. U.S.D.A.Technical Bulletin No. 4,14 p.

203 Sansone,C., T. Isakeit, R. Lemon, and B. Wanick. 2002. Texas Cotton production, EmphasizingIntegrated Pest Management.Texas cooperative Extension service Bulletin8-6116., p. 89. Schwartz,M. D., andR. G. Foottit. 1998.Revision of the NeararticSpecies of the-Genus Iygus Hahn,with Reviewofthe PalearticSpecies (Heteroptera: Miridae). Associated Publishers.Gainsville FL. Sevacherian,V., and V.M. Stern. 1972. Sequentialsampling plans for Iygzs bugs in califomia cottonfields. LLygushesperus, Lygus e/rsns]. Environ. Entomol. l:704- 7r0. Snodgrass,G. L., w. P. Scott,and J. w. Smith.1984. Host plants of raylorilyguspallidulus ard,Polymerus basalis (Hemiptera: Miridae) in the Deltaof Arkansas,Louisiana, and Mississippi.Fla. Entomol.67 : 402-408. stitt, L.L. I 949.Host plant sources of Lygusspp. infesting the alfalfa seed crop in soutlern Arizonaand southeastern Califomia. J. Econ.Entomol. 42:93-99. Womack, C.L., and M.F. Schuster.1987. Host plants of the tarnishedplant bug (Heteroptera:Miridae) in thenorthern blackland prairies of Texas.Environ. Entomol. 16:1266-1272. Young,O. P. 1986.Host plantsof the tarnishedplant bug, Lygus lineolaris (Heteroptera: Miridae)Ann. Entomol.Soc. Amer. 79:747-762.

204 vol.28 NO.3 SOUTHWESTERNENTOMOLOGIST SEPT.2OO3

ARTHROPODSPRESENT ON REMOVED FOLIAGE FROM AN APPLEPACKING LINE

JamesD. Hansenr.Laura R. Lewis2'3.and Gilbert F. Simmonsr'a

ABSTRACT

The effectivenessof packing houseoperations in removing artluopodson leaveswas evaluatedby surveyingfour applecultivars, Malus domesficBorkh., from 17 grower lots over six months of inspection. Among the 6y'00 leaves inspected,damage from the westem tentiform leafrniner,Phyllonorycter elmaellaDoganlar & Mutuura(Lepidoptera: Gracillariidae) was seenin 63Yoof the leaves. Twospottedspider mites, Tetranychusurticae Koch (Acari: Tetranychidae), found on 62 leaves, and predatory mites, Typhlodromus spp. (Acari: Phytoseiidae),found on 34 leaves,were the most frequentlive arthropods.Six other live artlropods (tlree parasiticwasps, a psocid,an aphid, and a spider)were also observed. Only one live arthropod,a predatormite, was found at the final pack samplingstation. The numbers of arthropodsdecreased with fruit processingalong the packing line. Data from this study indicate that foliage packedwith fruits is not a likely pathwayfor spreadinglive artluopods.

INTRODUCTION

Expandedglobal tade hasinueased the opportunitiesofexotic peststo expandinto new territories(LaGasa et al. 1997). The PacificNorthwest is a major exporterof domesticapples, Malus domesticBorkh. Fromthe 2000harvest, about 5 x lOsM.T. (6.5 x 105tons) of apples from WashingtonState were exported(NHC 2002). Becauseof this largevolume, there is concern that foliage included with the packed fruits may be a potential source for the infoduction of new pests. Knowledgeof the numbersof arthropodssurviving the packing line is essentialfor establishingpest risk assessment(Griffin 2001). Currently, only Mexico limits the amount of foliage at two leavesper carton(Erikson et al. 1996),and only Japanrequires applesto be fumigatedas condition for entry (Hansenet al. 2000). However,future restrictions from other countriesare possiblebecause ofexpanding world trade. Previous studies on the occurence of arthropod pests in packed fee fruits have emphasizedthe fruits themselvesrather than exhaneousplant material suchas leaves(Curtis et al. 1992, Johnsonet al. 2000, Knight and Moffrtt 1991, Moffitt 1990, Hansenand Schievelbein2002). Yet, the risk of inhoducingexotic pestsfrom leavespacked in fruit containershas not beenevaluated. A thoroughinvestigation would involve inspectingleaves

t USDA-ARSYARL, 5230Konnowac Pass Road, Wapato, WA 98951 2 Collegeof Agriculture, WashingtonState University, Pullman,WA 99164 3 Currentaddress: Departmentof EnvironmentalDesign, Univ. Califomia, Davis, CAg66l6 aCurrent address: 1460 Armstrong Ave., Clovis,CA 9361I

205 removedat points throughoutthe conventionalpacking process, beginning with the arrival of freshly harvestedfruits and examiningthe foliage for the presenceof arthropodsas evidence of phytosanitation. Thus, the objective of our study was to observediscarded leaf material along a typical packing line to determinewhere arthropods are eliminated.

FIG. 1. The arrangementof sampling stations along the packing line, each with an identifi cation codenumber.

MATERIALS AND METHODS

Apple leavesfrom four cultivars('Gal4' 'Fuji,' 'Delicious,'and'Golden Delicious') were obtainedfrom l7 growerlots (fruit harvestedon a particular datefrom one orchard)over six months.Foliage samples were collected from tenstations during packing (Fig. l), with the majority of the samplesconsisting of 'Delicious' leaves.The amountsvaried by growerlot, but a minimum of 20 leavesper lot per site werecollected in mostcases, Regardless, all leaf materialwas supposeto be removedbefore fruits werepacked. In the pre-sortsamples, foliage was obtained while fruits were still in bins and from drenchtanks. Some of the fruit was packed immediately, whereasothers were placed in conventional refrigerated storage or contolled atrnospherestorage to be packedlater. Sampleswere obtainedbetween pre-sort and the sort table from fruits in rinsetanks, from leaf trapson tank drains,and from fruits along the packingline. Foliagesamples were brought to the laboratorywhere they were inspectedunder a 30x binocularmicroscope. Arthropods were counted, removed, identified, and their viability determined. Data were analyzedusing SAS (SAS Institute1989). PROCMEANS was usedfor univariatestatistics. Becauseof heterogeneityin variances,nonparametic tests were usedto determine significant differencesby first arrangingdata by PROC RANK, then performing PROC GLM. This is the equivalentto a Mlcoxon rank sum test for two samplesand the Kruskal-Wallis &-sampletest for more than two samples. 206 RESULTSAND DISCUSSION

The total numberof leavesexamined was 6,408 among336 samples.The quantity of leavescollected varied among the samplingsites because of differential removal. Damageby the westemtentiform leaftniner,Phyllonorycter elmaellaDoganlar & Mutuura (Lepidoptera: Gracillariidae)was the mostprevalen! occurringin about63% ofthe leaves.However, because this insect is not a pest on fruits, it was excludedfrom any further analysis. only ten leaves werecollected at Station3, andthese data were also excluded. The largestgroup of live arthropodsobserved on leaveswas the twospottedspider mite, Tetranychusurticae Koch (Acari: Tetanychidae);there were 62 leaveswith spidermites, which were generally collected at Stations7, 8, and 10. The next most abundantgroup was the predatorymites, Typhlodromnrspp. (Acari: Phytoseiidae),recovered from 34 leavesalso at the samestations. The greatestnumber of live insectscollected tluoughout the study were tluee pamsiticwasps, although a live psocopteran,and an aphidwere found at Station g. Among all cultivars, the only live arthropodfound at the final pack (Station6) was a predaciousmite.

1400 1200

a () 1000 cc () 800 F1 600 ,ci z 400 200 0 8 910 Station

'Delicious' FIG. 2. Totalnumber of leavescollected at eachof the samplingstations along the fruit packing line.

'Delicious' Among the cultivarsexamined, only had leaf samplescollected at each station(Fig. 2). The percentageof clean'Delicious' leaveswas significantly different among the samplestations (F:3.21; df : 8, 175;p < 0.01). Significantdifference in ttre percentage 'Golden of cleanleaves was also found for Delicious' amongStation 4 through 7 (F= 3.g4;af

207 : 3, 55;P < 0.05).The cleanest station for boththese cultivars was Station 6 ('Delicious':mean 'Golden + SEM : 95.5+ 1.9%; Delicious':mean i SEM - 89.1+ 3.5%). Therewere no significantdifferences among the samplingstations for the remainingcultivars. Station 7 was the only samplingstation that had a significantdifference (F= 9.72;df : 2,55;p < 0.01) in percentageof cleanleaves per sampleamong cultivars ('Delicious': meanr sEM : 70.6+ 3.4%;'GoldenDelicious':mean+SEM:57.5+7.8%;'Fuji':mean*SEM=94.5+2.1o.), perhapsdue to the different growing conditions. Overall, 'Delicious' had the leastpercentage of cleanleaves among sample stations (mean * SEM= 80.4*3.lyo),but this may becausethe sampleswere collected from morestations than for anyother cultivar.

0.25

0.20 -9 a E cc 0.15 V) C) 0.10

0.05

0.00 8 910 Station

FIG. 3. Mean (+ SEM) number of live spider mites per sampling station for'Delicious' leaves.

If cleaning was effective along the packing line, then the number of living pests should be reduced as the fruits are processed. To test this, we examined the live infestation rate of the 'Delicious' most populous pest, spider mites, on among the sampling stations and found that the infestation rate decreasedrapidly after Station 7 from a mean (* SEIO = 0.22 (+ 0.18) mites/sampleto a mean (+ SEM) = 0.01 (+ 0.01) mites/sampleat Station 9 (Fig. 3). Samples from the sorting stations averagedless than 0.01 mites/sample. Hence, either live mites were eliminated early in the process or they died by the time they reached the packing table. The only mites, including the European red mite, Panonychus ulrni (Kock) (Acari: Tetranychidae), and predacious mites, collected at Station 6 were dead. The findings of our study provides support for the systems approach (Jang and Moffrtt 1994, Moffitt 1990, Moffrtt 1997) to quarantinesecurity. In the systemsapproach, production components, starting with field control and ending with fruit cleaning and sorting in the packing house, accumulatively assure a pest-free product. Besides being efficacious, other proven 208 advantagesof the systemsapproach are maintenanceof fruit quality without additional treatment facilities, reducedlabor costs without compromisingwork safety, and reduced harmful effectsto tlre environmentthrough deceases in wastedisposal. No firther treatment may be neededto preventpotential spreadofarthropod pestsfrom foliage including.thosein the containersholding the packedfruits. To assesspest risk, Griffrn (2001)recommended determining initial pestinfestation levels as the fust step. In our study, very few live arthropodsreach the packing house, regardlessofthe typesofapples harvested.Those that went throughthe cleaning,sorting and packingoperations did not survive. Ifthis studyis characteristicof otherconventional packing houses,foliage material that may be inadvertentlyincluded with the packedfruits very likely doesnot harbor any live arthropods.

ACKNOWLEDGMENTS

We thankM. Watkins,M. Heidt,D. Albano,K.Zapel,M. Weishaar,and B. Humphrey (ARS, Wapato,WA) for their assistance;P. Landolt(ARS, Wapato, WA), J. Johnson(ARS, Parlier,CA),C. Daniels(WSU, Richland,WA), and T. Gorrebeeck(Sarah Lawrence College, Bronxville, NY) for reviewingthe manuscript; J. Archer and F. Scarlett(NFE, Yakima, WA) for their contributions;and the anonymousgrowers and packinghouse for their cooperation. Supportfor this researchwas provided by the WashingtonTree Fruit ResearchCommission.

LITERATURE CITED

Curtis,C. E., J. D. Clark, J. S. Tebbets,and B, E. Mackey. 1992. Incidenceof arthropods fotrnd in packednectarine fruit in centralCalifomia. Southwest.Entomol. 17:29-39. Erikson, K. A., D. S. Brasch,and K. L. Casavant.1996. Washingtonexpansion into the Mexican fresh fruits and vegetable market: Washington apple competition and transportationalternatives. Wash. St. Univ. IMPACT Center Inform. Ser. #87. Pullman,Wash. 49 pp. Griffin, R. L. 2001. Pestrisk analysis:an altemativefor methyl bromide. North Amer. Plant. Protect.Org. Bull. 16:18-21. Hansen,J. D., S.R. Drake,H. R. Moffitt, J. L. Robertson,D. J.Albano, and M. L. Heidt. 2000, A two-componentquarantine teatment for postharvestcontrol of codlingmoth on apple cultivarsintended for exportto Japanand Korea. HortTechnologyl0: 186-194. Hansen,J. D., and S. Schievelbein.2002. Apple samplingin packinghouses supports the systemsapproach for quarantinecontrol of codlingmoth. Southwest.Entomol. 27 t 277- 282. Jang,E. B., and H. R. Moffitt. 1994. Systemsapproaches to achievingquarantine security, pp. 225-237./z; Sharp,J.L. Hallman,G.J. (eds.), Quarantine Treatments for Pestsof Food Plants. WestviewPress, Boulder, Colorado. Johnson,J, A., K. A. Valero,M. M. Hannel,and R. F. Gill. 2000. Seasonaloccunence of posthawestdried fruit insectsand their parasitoidsin a culled fig warehouse.J. Econ. Entomol.93: 1380-1390. LaGasa,E., J. D. Bell, andv. Mastro. 1997. Pathwaysfor exoticintroduced species. North Amer.Plant. Protect. Org. Bull. 15:18. Knight,A., andH. Moffitt. 1991.Removal of codlingmoth injured apples in packinghouses. Proc.Wash. Hort. Assoc.86l-2ll-213. Moffitt, H. R. 1990. A systemsapproach to meetingquarantine requirements for insect pests of deciduousfruits. Proc.Wash. Hort. Assoc.85: 223-225.

209 Moffrtt, H. R. 1997. A systems approachto meet arthropod-basedexport quarantine requirementsfor apple.North Amer.Plant. Protect. Org. Bull. 15: 35. NHC (NorthwestHorticultural Council). 2002. Apple fact sheet.www.nwhort.org/applefacts. htn. SASInstitute. 1989. sAS/srAT user'sguide, release 6.I I edition. sAS InstituteInc., cary, NC.

2r0 vol.28 NO.3 SOUTHWESTERNENTOMOLOGIST SEPT.2003

EFFECTOF SPINOSADON HONEY BEES(HYMENoPTERA; APIDAE) IN GUATEMALA

JohnP. Spencer,Jorge Ibana, pedro A. Rend6n

USDA-APHIS-ISMedfly Region US Embassy-Guatemala,Unit 3319 APO AA 34024_33t9

ABSTRACT

A field evaluationof the effectsof aerially appliedspinosad bait spray (Success 0.02.CBrM),used by the area-wideMediterranean fruiffly, Ceiatitis capitata'(\iiedemann) eradication program in Guatemala,on honey bees,Aphis melliferali-"*r, and hive activity was conductedduring the 2000 spray season. The parametersmonitored were adultbee mortality, brood size,pollen reserves, panels ofhoney, panelsofadults, andhive weight. Statisticalanalysis of the data revealedno significant differencesbetween parameters in the treatedand control area,demonskating ihat there were no deleterious effectson thebee population due to the spinosadapplications.

RESUMEN

Se efectu6 una evaluaci6nen campo del efecto producido por el insecticida spinosad con cebo (Success0.02 cBfr), aplicadoen aerosol,sobre a'bejasadultas y la actividad de sus colmenas durante la temporada de fumigaci6n del afro 2000. Este compuestoquimico se utiliza en 6reastratadas por el prograriade erradicaci6nde mosca del mediterr6neo. Los par6metrosmonitoreados fuiron: mortalidad de adultos, cria operculada, reservasde polen,panales de miel, panalesde adultosy pesode la colmena.El an6lisis estadisticode los datos no indica una diferenciasignifrcativa eniie .t 6rea de hatamiento y el irea de control, demostrandoque no exisie un efecto nocivo en la poblaci6nde abejasdebido a la aplicaci6nde spinojad.

INTRODUCTION

The Meditenanean fluit fly,.Ceratitiscapitata (Wiedemann), commonly known as Medfly, is one of the most destructivephytophagous insect pestsin the worlJ, with larvae attackingmore than 300 speciesoffruits andvegetables (Liquido ,t ur- rsgli Becauseof its destructive nature,large-scale conlrol p.ogtutir havebeen implemented in variousparts of the world including Mexico andGriatema=la. Medfly was firsi detectedin Guatemalain 1976 and in Mexico in 1977. Thesedetections led to a major effort that resultedin its eradication from Mexico in 1982and the establishmentof a bairierzone in cuatemala. The chemical componentof the eradicationprogram consisted of a mixture of malathionand hydrolyzedprotein that was integratedwiih the releaseof sterileflies and the establishment of an effectivequarantine ptof*. The banier preventedthe northward movementof Medfly back into Mexico until 1999, at which time 274 outbreakswere 211 detectedin southernMexico. In 1999, emergencyfunds were appropriatedto eradicate Medfly from Mexico and Guatemala,and planswere madeto begin the aerial spraying phaseofthe programin February2000. The chemicalproduct selected for this phaseof the Guatemalaprogram was Success0.02 CBrM produced by DowAgroSciences (Indianapolis,IN), This productcontains spinosad, a mixture of spinosynA and spinosyn D, derived from the fermentationof the actinomyceteSaccharopolyspora spizosa, which has improvedenvironmental properties over older insecticides(Cleaveland et al. 2001). Success0.02 CBt" bait is reportedto be an improvedbait combininglow userates and a reducedrisk toxicantto non-targetinsects (Anonymous 2003); however, its potentialeffect on honey bee, Aphis mellifera Linneaus,was an obvious concern since one of the componentsis sugar. In I 999,prior to recommendingthe use of Success0.02 CB* as a componentof the eradicationprogram, Guatemala APHIS PPQMethods Development conducted a seriesof field cageevaluations, as well as a large field evaluationin southwesternGuatemala, to determinethe effectsof Success0.02 CBrMon honeybees. The conclusionsfrom these evaluationswere that Success0.02 CBrMwas not toxic to beesand that the product had no deleteriouseffects on normalhive activitywhen used according to labeldirections (Rend6n et al., 2000). In February2000, the GuatemalanMinistry of Agriculture,Ministerio de Agricultura, Ganaderiay Alimentaci6n (MAGA), appointeda commissionof MAGA personnelto conductanother field evaluationon the effectsof Success0.02 CB* on honey beesin collaborationwith USDA-APHIS-ISand Moscamed (Spanish for Medfly) program personnel. This evaluationwas done in conjunctionwith the Medfly area-widespray progrzrmto evaluatethe effects from actual field use of repeatedapplications across multiplegeographies and climates.

MATERIALS AND METHODS

The evaluationwas conductedfrom 16 Februaryto 2 April 2000' Apiarieslocated insidethe sprayzone were nearthe villagesof SomberitoBajo and CaserioLos Perezof the municipalityNuevo Progreso, San Marcos, Guatemala; apiaries outside the sprayzone werebetween San Antonio Las Floresand Palin, also in NuevoProgreso, San Marcos. The areais 400-600m abovesea level; coffeeis the predominantagricultural crop. Sixteenbee hives from five apiarieswere selectedwithin the treatmentarea and I I hives from four apiarieswere selected outside the treatmentarea to seryeas controls. Controlapiaries were 1.5-5.5km from the spray zone. Prior to the initiation of the evaluation,hives were examinedto determinetheir generalcondition and absenceof disease. Two aspectsof honeybee activity were evaluatedin order to determinethe potential effect of Success0.02 CBfr; the first aspectconsidered adult mortality aroundthe openingof the hive while the secondconsidered the effect ofthe productupon hive vitality. In orderto determineadult bee mortality, apron traps (Rhodesand Wilson 1978)were constructedimmediately in front of the hives. The aprontraps were constructed by placinga white cloth on the ground and surroundingit with concrete-blocks 20 cm high, creatingan area lxl m. Deadbees falling onto the cloth in the 1-m2area were countedand removeddaily. The effect of Success0.02 CBru upon colony activitieswas determinedby taking measurementsof brood(square centimeters), pollen reserves (square centimeters), hoirey (number ofpanels), andoverall hive weight. Initially, measurementswere taken weekly, but the lasttwo series of measurementswere taken at two-week intervals. The Success0.02 CBil concentratewas mixed with water (3 pats water:2 parts Success)to producean 80 ppm activeingredient mixture which was aeriallyapplied to the treatmentarea using LET-410 and Turbo Thnrshfixed-wing aircraft at the rate of 0.38 g a.i./ha.Aerial applicationswere made at weeklyintervals for a total of sevenapplications.

212 In order to comparethe number of dead bees from hives inside and outside the spray zone, the Wilcoxon test for paired sampleswas used. The non-parametricMann- Whihey test for two independentsamples was used for comparinghive activity variables (InfanteandZarate de Lara 1996).

RESULTSAND DISCUSSION

Honeybee mortality is illustratedin Fig. 1. The daily meansof beemortality were 20 and 13 for the hives locatedin the control and treatedareas, respectively. Statistical analysisrevealed no significantdifference between the two areas(Z= 1.8781;p-value: 0.06M).

q) .l 250

J & Eoo , a 3 r5o

€ 100

*E nt X.o

Weeks

FIG l. Honeybee mortality from areatreated with Success0.02CBrM vs non-treatedarea in Guatemala

The effectsof Success0.02 CBfr on hive activity is presentedin Table l, Only during the first week were pollen reservesand panelsof adults significantly higher in the control hives than the hives in the treatedarea. The only other significant differenceswere nearthe end of the evaluationin which panelsof adults,brood size,and pollen reserves were significantly higher in the treatedarea than the untreatedarea. This is counterto the hypothesisthat the applicationsof Success0.02 CBrM may havea deleteriouseffect on hive viability. Results of this evaluationare consistentwith the findings of the evaluation previouslydone by Rend6net al. (2000). In that study,there was no beemortality caused by the aerial applicationsof Success0.02 cBn; nor was there any negativeimpact upon normal hive activity as measuredby brood size,pollen reseryes,panels of honey,panels of adults,and hive weight. Burnset al. (2001)also reported no effectson hive conditionsor brood from spinosadbait spray applied to commercialcitrus to control Med fly and Caribbeanfruit fly, Anastrephasuspensa (Loew) in Florida. The susceptibility of bees to direct application of spinosadhas been reportedby variousinvestigators. Mayes et al. (2003)have reviewed much of this work. The absence of acuteadult mortality and lack of deleteriouseffects upon hive activity in this studyis most likely the result of a combinationof factors. One is the avoidanceof the Success 0.02cBr'mixture by the bees due to the presenceof l% ammoniumacetate. The 213 6? c.l$r|c..lo ,\zt .!j (o\\o\oF- vtr r'- -: c.i ri \o =la ca s l € -l a) C-\@C-.1 et v?-\F) el E 9 at o S€C.lO\\o ci6$c.tm *t r o

LI ot - a =os\oF- C) Or)66 ztx :-iooir- I bl Et _ trt Y A h bc\o\8.F- dlx n-1-'19q al a{ a

o a)l \O t'- O\ € coc-Oc{O\ >r 3E 11 .9. .n@

j-t C) I c.l cO t.- \O \O ? *t = q) >{ i) \q.1v'ln $a ol ,E !\Og$ !) o a ol llI (n C) d

f,] _ o oo$o6.to\ F ct v v't9qflq o ONF-sfF- O arl 6 OO\F- PI O aliv q)l q) ol ol + F g trlE Di .o trt (! b s € H () ot o r; +-i r-od o1 OF.-6\0 c.l A

oO C.l t-- o o r o ndlv.lbvl SOri$ (/) o\ci6\0 el 6 Or)o6oo @ ol L) m aq c.l (\l q NI (/)l a'.1 a.l @ F- F xl q) ne!-'lcJ\ F- c.r $ e.l trl61 q)6 \o!\oF- -t \t\OooOO\ F cn c..l c.l c.t c-l €!) I .l I q) F C) ..1 c.r Yf \ F B

214 repellencyofthis chemicalcompound to honeybees has been well documentedby Tarshis Moreno(2001). Althoughthe inclusionof ammoniumacetate into the Success0.02CBrM mixture attractsMedfly andAnastrepha spp. which readily feed upon the mixture to obtain the protein necessaryfor egg development,bees wili avoid the mixture given the opportunity to do so, as in a field cageor field situation(Rendon et al. 2000). Edwardset reportedsignificant {a(2003) honeybee mortality when exposed to GF-120lSuccess 0.02 cB9 fruit fly bait in the laboratory. However in that expoiure study, beeswere confined petri to dishes where their normal avoidancebehavior was negate; and drrect contact unavoidablebecause ofthe restrictedspace. other importantfactors leading to the absenceof any effectswere tle low dosage rate and low volume of the aerial applications(droplet size of 3000-6000microns in diameter dispersedat the rateof 60-80droplets per m2j. The probabilityof a honeybee in flight receiving enough spinosadto causeaeatrr is very remote. Additionaily, through a very active environmentalmonitoring program in Guatemala,chemical analyses of water andvegetation sampleshave verified that residues of spinosadin the field arebroLen down rapidly and do not accumulate(USDA-APHIS-IS 2000, USDA-APHIS-IS 2001). These findings are consistentwith residuepersistence studies done by saunders-a n..t who ltsez; foundthat spinosaddegraded rapidly in the field dueto photodegradationlysunlight. - Tlese investigatorsare in agreementwith the positiontakeniy stark et al. (1t95) that.predictions regardingthe safety of insecticides1o beneficialinvertebrates must be validated with field data. Thesedata indicatethat under the conditionsthat the Medfly area-wide eradicationprogram currently operates,no detrimentaleffects to beesor honey production haveoccurred or arelikely to occurfrom applicationsofSuccess 0.02 cBil.

LITERATURECITED Anonymous' 2003. GF-120 Naturalyte Fruit Fry Bait Technical Bulretin. Dow AgroSciences,Indianapolis, IN. Burns,R. 8., D. L. Hanis,D. S. Moreno,and J. E. Eger. 200r. Efficacyof spinosadbait spraysto control Mediterraneanand Caribbeanfruit flies (Dipteia:Tephritidae)in commercialciffus in Florida.Florida Entomol. g4:672_67g. cleaveland, 8., c. M. A. Mayes,and S. A. cryer. 2001. An ecologicalrisk assessmentfor spinosaduse in cotton.pest Manag Sci. 5g:70_g4. Edwards,c. R', c, K. Gerber,and G. J. Hunt. 2003. A laboratorystudy to evaluatethe toxicity of the Meditenaneanfruit fly, ceratitis capitata,bait, success0.02 cB, to th3honey bee, Aphis mellifera. Apidologie. 34:17l-1g0. Infante,G. Saidy p. Guillermo Zarateai rara. 1996. M6todosEstadisticos. un enfoque interdisciplinaio. 2a. Ed. Mexico, Trillas, Tercerareimpresi6n, eurii. rags. s:z- 554. Mayes, M' A., G. D. Thompson,B., and M. M. Miles. 2003. spinosadtoxicity pollinators to andassociated risk. Rev.Environ. contam. roxicor.-iis:31_og. tn p..rr. Liquido, N.J., L.A. Shinoda, and R. T. Cunningham, 1991. Hori-pt*t, of tf,. Mediterranean fruit f,lr (Diptera:Tephritidae):anannotated world review.Entomol. Soc.Amer. Misc. pub. No. 77.52 pp. Rend6n,P.A., F. Jer6nimo, J. Ibarra,uttd v.c. Arvarez. 2000. Effectivenessof Success 0.02cBru for the control of fruit flies and its effect "-u""r-,grJ meilifera L. USDA/APHIS/ppe, MethodsDevelopment Station, Cuute_uiu. fb-pp. Rhodes,H.A., andw,T.yt:gl. 1978. A simpletrap for..r";;;;;d# adulthoney bees.Amer. Bee J.118:671-672. saunder,D.G., and B.L. Bret. 1997. Fateof spinosadin the environment.Down to Earth. 52:14-20.

2t5 Stark,J.D., P.C. Jepson,and D.F. Mayer. 1995. Limitationto useof topicaltoxicity data for predictionsofpesticide side effects in the field. J. Econ.Entomol. 88:1081- 1087. TarshisMoreno, A.M. 2001. Detectionof the phototoxicdye phloxineB in Anastrepha ludens (Loew) (Diptera:Tephritidae),Apis nellifera L. (Hymenoptera:Apidae)and honey.M.S, Thesis.University of Texas-PanAmerican, Edinburg, Texas.49 pp. U.S. Departmentof Agriculture,Animal andPlant Health Inspection Service, International Services. 2000. EnvironmentalMonitoring Report. SpinosadSpray Trial for MoscamedEradication Program in Guatemala.Intemal Report. 36 pp. U. S. Departmentof Agriculture,Animal andPlant Health Inspection Service, International Services. 2001. EnvironmentalMonitoring Report. SpinosadSpray Trial for MoscamedEradication Program in Guatemala.Internal Report. 49 pp.

216 vol.28 NO.3 SOUTHWESTERNENTOMOLOGIST SEPT.2OO3

ARTIFICTALSI.JBSTRATES FOR OVIPOSITIONOF ffOTOIlE CTA IRROMTAUHLER (HEMIPTERA:NOTONECTIDAE).

c. solfs-Rojas,H. Quiroz-MartInez"y. A..Rodrfguez-castro, M. H. Badii,A. E. Flores,L. J. Gal6n-Wong',R. Foroughbackhch

Laboratoriode Entomologfa, Facultad de cienciasBiol6gicas, U.A.N.L.; Apdo postal 105- F, ciudadUniversitaria, c,P. 66450,san Nicol6s de los Garza,Nuevo Le6n, M€xico. e-mail:[email protected]

A subsrateand depth r**."f:ff$l caniedout with themosquito rarvar predatorNotonecta iftoruta uhler in artificial pools from March to December1996. Four artificial substrates(cotlon thread,nylon thrad, Agave sp. fibers and tree twigs) were tested at threedepths (10, 20 and30 cm) as ovipositionsite bynotonectid female. {esults showed two peaksin eggproduction and preference for Agave sp. fibers and the 30-cmdepth level as ovipositionalsites.

INTRODUCTION

Notonectisare voraciouspredators of mosquitolarvae and are thereforeconsidered promisinggroup for biologicalcontrol (Garcia 1983, Garcia and Des Rochersl9g4). The qredatorycapacities of severalspecies of backswimmershas been reported previously @llis & Borden1970, Toth and chew 1972,Gittelman 1974, Fox and Murdoch 197g,ihesson 1984). Miura and Takatrashi(personal communication)released Notonecta unilasciata Guerin eggs to control Culex tarsalis Coquillet in rice field with excetlentresults. Neri- Barbosaet al' (1977) releasedadults of .lVoronectaitorata Uhler in artificial containersto control larvae of Aedes aegtpti (L.) and culex pipiens L.; Their results showed low mosquitolarval densitiesand notonectidnumbers after ten months.However. cost of colonizatio:tand massproduction of thesenotonectid species, coupled with the logisticsof their distribution,handling ard timing of releaseat the appropriatebreeding site have impededtheir use as mosquitocontrol agents(Legner 1993). vtiura (19g6),ind Legner (1995) have suggestedinoculating mosquito breedingareas with backswimmereggs collected on artificial oviposition material. The objective of this study, therefore,was to determinethe preferenceof Notonectairrorata to different artificial substratesand substrate depthlevels as ovipositionsites. This informationmay useful as a supporttool for integated managementof mosquitoesin environmentsinherent to Mexico.

MATERIAL AND METHODS

The studywas carriedout from March to Decemberof 1996at the Agricultural Field Station of Instituto Tecnol6gicoy de Estudios superioresde Monterrey ltresu; in trre munrcipalityof Apodac4Nuevo Le6n. Two concretepools, each gxgxlm, with a continuous

iD"p Dffiicr',c66iogi;ehmunologla, F.c,B., U.A.N.L,. A.p. zlgo.Monterrey, N.L. Mdxico c.p. 66450

217 water supply were used in the study. Submergedvegetation, with Chara sp. being the dominantplant, was presentin the boftomof thesepools at all times.The watertemperature in the pools rangedbetween 20 to 39oC and pH variedfrom 7,5 to 8.1. The pools were continuouslyexposed to solarradiation. The artificial substratesexamined as oviposition sites for Notonectairrorata were 30-cmstrips ofcotton thread,nylon threads,Agave sp. fibers,and tree twigs. substrateswere suspendedfrom the surfacewater by red threadattached to 5x2.54-cmpolystyrene squares at thrce lO-cm.depth levels denoted as level 1,2 and3.respectively. A plummetwas attached to each srip in order to keep it vertical. There were four replicatesfor each substrate. substrateswere changedevery 15 days, and the numbersof notonectidseggs on each substrateat eachdepth level were recorded.Analysis of variancewas usedto determine signifrcancein ovipositionpreference for substratesand depth level according to Zar (1974).

RESULTSAND DISCUSSION

Notonectairrorata eggswere first observedon the artificial substratesapproximately 6 monthsinto the 8 monthstudy (i.e., 22May 1996)(Fig. 1). The highestnumber of eggs recorderoccuned in late spring(22May) with ll7 eggsand early summer (26 June)with 125eggs as shownin Fig. l. Numberof eggsdeoeased on the remainingsample dates. This pattemof ovipositionand the relativelow numberof eggssuggest that, during these months, the backswimmersprobably did not have the appropriatequantity and quality of food. Rodriguez-Castro(personal communication) observed that, under laboratory conditions with appropriatefood such as mosquito larvae,lf. irrorata urd Bueonascimitra produceeggs continuouslythrough the year. Backswimmeroviposition rates varied in with the differentartificial substratestested. The most attractive material was the Agave sp. fibers (Fig. 2). Difference in oviposition substratepreference was consistentthroughout the study.As canbe notedon Fig. 2, the total numberof eggsdeposited on the differentsubstrates was Agave sp. fibers,(310); nylon fibers(61); tree twigs, (31) andcotton thread (6). Statisticaldifferences were found between thesesubstrate preferences (P < 0.05).We thusconclude that the backswimmerexhibited a preferencefor Agavesp.fibers as an ovipositionsubstrate. This may be due to it beingthe materialmost similar to thatof naturalsites where females lay eggs.

140 o 912o E g loo (/)80

{on ovv o Eao Ero F 0 o oo GO =EEEEE9O o ==

FIG. l. Total number of eggs deposited by Notonecta irrorata females on all oviposition substratesduring the period March to November 1996 in artificial pools at the Agricultural Field Station of ITESM, Apodaca,Nuevo Le6n, Mdxico.

2r8 $ssooo FE Nts$ ++ SN ON fif;iisrtg$gHE Date

FIG. 2. Substratepreference by ovipositing Notorocta irrorata femalesduring the period March to November1996 in artificial pools at the AgriculturalField Stationof ITESM, Apodaca,Nuevo Le6n, M6xico.

70 60

o50 o, ctt uJ 40 b b30 It Ezo z 10 0 99&STFB *F+Fg $$ OOYYZ NF R..:N*frfifiiEic{ EsSeg Date llG. l. D"pth levelpreference by ovipositing Notonectaitorata femalesduring the period March to November1996 in the AgriculturalField station of ITESM, Apod-aca,Nuevo Le6n,M6xico.

Femalesof N. inorata usuallyselected the deepest portion of thestrips located 30 cm below water surfacefor oviposition.A statisticaldifference was found amongthese depth level results(P < 0.05). As shownFig. 3, the numberof eggsat eachdepth was 49 eggsfor the first level, 145eggs for the secondlevel, and 214 for third level.This studyprovides a baselineof informationabout ,iV irroruta andthe possibilityof massrearing this notonectid speciesunder artificial conditions,and a methodto manipulatethis aquaticpredators. This informationcan improvethe useof backswimmersas a biologicalcontrot of mosquitolarvae.

2r9 ACKNOWLEDGMENT

study was supporte!by consejo Nacional . TSt"Esfiategias de ciencia y Tecnologiathrough the p19j..t de lvlanejoIntegrado de Larvasde Mosquitos;coNAc;i Ref. 0651p- M9506'The AgriculturalField Stationof InstitutoTecnologico y de EstudiosSuperiores of MonleneyITESM, locatedin Apodaca,Nuevo Le6n, M6xico. tjavid Lazcanovillarreal for translatingand critically reading drafts ofthe manuscript.

LITERATURE CITED chesson,J. 1984.Effeots of notonectids(Flemiptera: Notonectidae) on mosquitoes(Diptera: culicidae):Predation or selectiveoviposition. Environ. Entomol. l3(i):53 l-53g. Ellis' R' A. and J. H. Borden. 1970. predationby Notonectaundulin (Heteroptera: Notonectidae)on larvaeof yellow fever mosquito.Ann. Entomol. Soc. of Am. 63(4):963-972. Fox, L. R' and w. w. Murdoch. 1978.Effects of feedinghistory on short-termand long- term functionalresponse in Notorrectahofmanni. J. Anim. EcoL47t 945-949, Gallsuff,P. S.; F. E. Luja.-P. S. welch and J. G. Needman.1959. culture methodsfor invertebrateanimals. Dover publications. 590 pp. Garcfa, R. 1983. Mosquito Management:Ecological Approaches.Environmental Management.7(l):7 3-7 8. Garcfa,R. andB. DesRochers. 1984. studies of the developmentof an integratedmosquito conhol stratery for the Fall River Mills Area. SpecialFunds for the Mosquito Researchin Califomia.57-62pp. Gittelman,s. H. 1974.Manaiega hondurensisand Buenoa ontigone as predator of mosquito lalae in CostaRica (Hemiptera:Notonectidae). Pan-Pac Entomol. 50:84-85. kgner, E. F. 1995.Biological control of Dipteraof medicaland veterinary importance. J. VectorEcology 20( I ):59-I 20. Miura, T. 1986.Rice field mosquitostudies. University of Californi4 MosquitoControl research.Ann. Rept.pp 50-52 Neri-Barbosa,J. F, H. Quiroz-MartinezM. L. Rodriguez-Tovar,L. O. Tejadaand M. H. Badii. 1977. Use of Bactimos Briquets (8.1.i.) formulation combined with backswimmerNotonecta irrorata (Hemiptera:Notonectidae) for control of mosquito larvae.J. Am. Mosq.Control Assoc. l3(l):87-89. Toth, R. S. and R. W. Chew. 1972. Developmentand energeticof Notonecta undulata duringpredation on Culextarsalis. Ann. Entomol. Soc. of Am. 65(4):963-972. Zar,J.R. 1974.Bioestatistical analysis. Prentice-Hall, Inc. Englewood Cliffs. NJ.pp.

220 vol.28 NO.3 SOUTHWESTERNENTOMOLOGIST SEPT,2003 SCIENTIFICNOTE

FIRST RECORDSOF STETHORUSHISTRIO CHAZEAV (COLEOPTERA: COCCINELLIDAE)FROM THE TINITEDSTATES

DarrenA. Pollock' andGerald J. Michels,Jr.

TexasAgricultural ExperimentStation TexasA & M UniversitySystem 2301Experiment Station Road Bushland,TX790l2

Beetles of the coccinellid genus Stethorus Weise, with approximately 70 species worldwide, are obligate predators of spider mites (Acari:Tetranychidae)(Gordon and Chapin 1983). Many of these mites are important pests of various crops, fees, and omamentals.Because of this, severalspecies of Stethorushave been used in concerted effortsat biologicalcontrol of spidermites, especially Tetranychus urticae Koch. Someof thesecoccinellid specieshave been introduced,both intentionallyand unintentionallyin non-nativelocalities (e.g., Gordon and Chapin1983; Gordon and Anderson1979, Hoy and Smith 1982). Recently, an Austalian species,Stethorus nigripes Kapur, was collected in the Texas Panhandle in the United States (Gordon 1993). Its re-collection and its subsequentspread into adjacent Oklahoma and Kansaswas documentedby Pollock and Michels Q002). Our purposehere is to report the collection of Stethorushistrio Chazeatin Texas,representing the first authenticatedrecord of this speciesfrom North America. Accordingto Gordonand Anderson (1979), S. histrio was originally describedfrom La R€union,Mascarene Islands; and is now known to occur in Australia,New Caledonia andChile. Gordonand Anderson (1 979) suspected that the presenc€of S. histrio in Chile is dueto humanhansport, but thereis no uedible evidenceofthis. Gordonand Chapin (1983) expandedthe known rangeof S. histrioto includeMexico (Yucatan).According to Houston (1980),S. histrio is quite widespreadin Australia,being known from all states,including Tasmania.Specimens in the UnitedStates National Museum (N. Vandenberg,pers. comm.) are from Mexico, Paraguay,Chile, and Brazil. The Texas recordsreported here represent the first documentedpresence of S. histrio in North America"north of Mexico. Specimensof S. histrio were collected at severallocalities in the Texas Panhandle during August 2001, through routine sampling ofcom fields for speciesof Stethoru!; (see Pollock and Michels 2002 for a brief explanationof this project). Adults of S. histrio resemble closely those of S. nigripes Kapur, the latter of which was also recently documentedas being very commonin northernTexas and adjacentstates (Pollock and Michels2002). However,the larvaeof the fwo speciesare rather distinctive; mature larvae of S. nigripesare conspicuously pink in color, while thoseof S. histrio are gray. It was the discoveryof this differencethat led the authorsto suspectthe presenceof a speciesother ttnn S. nigripes. Adults, larvae and pupae of S. histrio were collected from two entirely different locations. A large redbudtree (Cercis canadensis,Fabaceae) in a yard in Amarillo (Randdl Co.), Texas, was moderatelyinfested with tetranychid mites, among which were found larvae and adults of S. histrio. Identification of this specieswas confirmed by

' Presentaddress: Department of Biolory, EasternNew Mexico University,Portales, NM 88130 22r dissectingrepresentative males and comparingthe genitalia with the diagnostic figures in Gordon and Anderson(1979). The larvae collectedwith the adultswere almost identical to thosefigured in Britton andLee (1972) ndGordon andAnderson (lg7g). The other situation in which specimensof S. histrio were found was in commercial com fields near the town of Hart, Texas (southemcastro co.). Routine samplingof comfieldsin the TexasPanhandle had been underway through 2000 and 2001, and we found a field nearHart that was very heavily infestedwith mites,and which had numerousadults andlarvae of Stethorus.The distinctivelygreyish larvae of S. histrio werequite common, as wereadults and laryaeof s. nigripes. Upon returnto the laboratory,we noiicedthat we had collectedthree different speciesof Stethorusfrom a singlecorn field: S. histrio, S. nigripes, andthe nativeS. caseyiGordon & Chapin. Prior to our researchin the TexasPanhandle, Stethorus diversity in the TexasHigh Plainswas thoughtto consistofthe nativespecies s. punctumpunctum ands. caseyi.Basid on the resultspresented in this paper and Pollock and Michels (2002), it is evidentthat Stethorusdiversity is poorly understood,and may be richer than expected. species of stethorus are well known from perennial habitats (e.g., orchards, omamentals),but little is known about speciesdiversity and ecology in annualhabitats. Given the ability of thesebeetles to rapidly colonizemite-infested annual habitats, it is reasonableto assumethat Stethorushas great potential as a biological control agent in annualcropping systems. conhol of spidermites in annualcrops, especially Banki grass mite in com, is increasinglydifficult as resistanceto acaracides,specifically bifenthrin (Capture@),is becomingcommon. Thereis a needfor further researchto determinethe number and range of speciesof Stethorusin the Texas Panhandle,and other parts of the Great Plains states. An understandingof the seasonalabundance, life history, and basic ecologyof speciesof Stethorusin annualcrops and perennialplant refugesis lacking and needsto be pursued. Voucher specimensof the larvaeand adults of all speciesof Stethoruscollected during the courseof this researchwere placedin the entomologycollection at the Texas AgriculturalExperiment Station Research and Extension Center at Bushland,TX. This project was funded in part by a grant from the Texas Higher Education coordinating Board Advanced ResearchProgram, Biological sciences ResearchArea (0227-reeg). LITERATURECITED

Britton, E,8., and B, Lee. 1972. Stethorusloxtoni sp. n. (Coleoptera:Coccinellidae) a newly-discoveredpredator of the two-spottedmite. J. AustralianEntomol. Soc. I l: 55- 60. Gordon,R.D. 1993.Stethorus nigripes Kapur new to North America,and a new synonymin StethorusWeise (Coleoptera: Coccinellidae). Southwest. Entomol. l8: 67-68. Gordon, R.D., and D.M. Anderson. 1979. The genus StethorusWeise (Coleoptera: Coccinellidae)in Chile. The ColeopteristsBulletin 33:61-67. Gordon,R.D., and E.A, Chapin.1983. A revisionof the New World speciesof Stethorus Weise(Coleoptera: Coccinellidae). Trans. Amer. Entomol. Soc. 109: 229-276. Houston,K.J. 1980.A revision of the Australianspecies of StethorusWeise (Coleoptera: Coccinellidae).J. AustralianEntomol, Soc, 19: 8l-91. Hoy, M.A., andK.B, Smith. 1982.Evaluation of Stethorusnigripes [Col.: Coccinellidae]for biologicalcontrol of spidermites in Califomiaalmond orchards. Entomophaga 27: 301- 310. Pollock, D.A,. and G.J. Michels, Jr. 2002. Distribution of Stethorusnigripes Kaptx (Coleoptera:Coccinellidae), a predatorof Banks grassmite foligonychuspratensis (Banks)lin the southemUnited States.Southwest. Entomol. 27:217-220.

222 F

vol.,28 NO.3 SOUTHWESTERNENTOMOLOGIST SEPT.2OO3 SCIENTIFICNOTE

EFFECTOF SUBSTRATEONPAECILOMYCES FUMOSOROT]SEUSI MYCOSIS OF HOUSEFLT LARVAE

GustavoA.Yiuqtez- Jaime,Sergio R. S6nchez-Pefia3/, Jer6nimo Landeros-Flores and EugenioGuenero-Rodriguez

Departamentode ParasitologiaAgricola, Universidad Aut6noma Agraria Antonio Narro, Saltillo,Coahuila, Mexico, C.P. 25315

Larvaeof housefly, Muscadomestica L., often appearto be resistantto topical fungal infections(e. g., Mullens etal 1987,Steinkraus et al. 1990,Geden et a't.1995, Leucona et al. 1996). However, inoculated larvae and pupae were not maintained in microbe-rich substratessuch as decayingorganic material similar to their natural habiiat in these studies, and this may be an important factor when trying to induce fungal infectionsin immaturehouse flies. Shieldset al. (1981)and Pereira et al. (1993)reported stronginhibition of Beauveriabassiana (Bals.) Vuill. in soil; however,this inhibitionwas rernovedafter subshatesterilization. Speciesof Aspergillus and,Penicilliutn are cornmon soil fungi and are stronginhibitors of Beauveria(Shields et al. 1981,Majchrowicz et al. 1990).Beauveria conidia in soil were more than seven orders of magnitudemore pathogenicto fire ants,Solenopsis invictaBuren, in sterileversus non-sterile soil (Pereira et al. 1993).Therefore, in this studyour objectivewas to determinethe effectofconidial concentation and incubation substates upon mycosis of house fly larvae. The fungal strain we utilized, identifiedas Paecilomycesfumosoroseas (Wize) Brown and Smith (Samson1974), was isolatedfrom five house fly pupae showing a densegranular, pinkish-brownmyceliar growth, collectedby G. A. V6zquez-Jaimein urban Saltillo, Coahuila.M6xico. To optimize fungal activity (Brownbridge et al. 2001), Paecilomyceswas inoculatedon harvesterants, Pogonomyrmex barbatus (Smith), and then isolatedfrom sporulatingant carcassesonto Sabouraudmaltose agar plus lYo yeastextract (SMAY). Cultures(second transfer on SMAY) were grown at 25-27"C for three weeks under fluorescentlight for conidiaproduction. House fly larvaefrom wild populationswere collectedfrom a fermentedbran paste(l kg of wheatbran mixed with 500 ml of distilled water) placed on the ground for three days in a swine and cattle bam with a large M. domesticapopulation. More than99% of flies thusreared from this bran(n=130) were M. domestica(Dodge 1958) @ 'Diptera:Muscidae 'Currentaddress: Molecular Genetics & Microbiology, I UniversityStation 45000, The Universityof Texasat Austin,Austin TX 78712-0162,USA. 'Corresponding author

223 Paecilomycesfumosorosew conidia were suspendedin 0.025% CenturyrM,an organosiliconesurfactant (union carbide, Danbury, CT) in sterile distilled water, and filteredthrough nylon mesh.Conidial viability, determinedmicroscopically fiom conidia incubatedfor 16 hours at 25"c on SMAY, was >98%. Third-instarhouse fly larvae, placedin a pieceof nylon mesh(10 lines/mm)were dipped for five secondsin 20 ml of conidialsuspensions in eachofeight concentrations(conidia/ml) as follows: 1.64x 108, 1.04x 10",1.64 x 70',1.64x l0o, 1.64x l0), and1.64 x l0a.Non-dipped larvae, and larvae dipped in water with surfactantonly were included as controls. Three groups of 13-23(mean=l9) larvae were exposed to eachconidial treatrnent, after which eachlarval grou,qwas incubatedseparately in 100-ml Petri dishesin one of three subsfrates:l) sterilizedmoist filter paper,2) a l-cm layer of moistenednon-sterile, fermented bran, or 3) a l-cm layer of non-sterilemoist cattle feces. Larvae were incubatedat room temperatureuntil pupation,after which pupaewere extracted and placed on moist sterile filter paperin Petri dishes;fungal or adult fly emergencewas recorded.Mortality values (as influenced by conidial concentration,incubation substrateand the two-factor interaction)were statistically analyzedusing the logisticregression model from SAS6.12 (sAS 19e6). Most teatments did not causesignificant mortality of housefly larvae or pupae. The maximummortality value (69%), and the only valuesignificantly different from the controls(P>0.021) occurred in pupaedipped (as larvae) at the highestconcentration (1.64 x 10"conidia/ml) incubated on sterilepaper. Also, visible.mycosisof pupaeresulted only at the threehigher spore concentrations, and only when larvaewere incubatedon sterile filter paper(5, 21, and 53% of pupaecovered with fungalsporulation). overall, conidial concentrationsand the interactionconcentration-substrate had a significant effect on mortality; overall, the effect of incubation substratewas not significant (p>0.021) probably due to only one concentration-substratecombination being so. These results, like previous reports, indicate that house fly larvae are relatively recalcitrantto fungal infections.Phenomena possibly causing this resistanceinclude the inhibitory effect of the larval substrate'smicrobial flora upon developmentof entomopathogenicfungi. when Pereiraet al. (1993)applied conidia directly to fire ants and transferredthem to sterile vs. non-sterilesoil, they did not observemortality differences.However, in the presentreport there was an effect of substrateon mortality regardlessofdirect applicationofspores to the insects'cuticle. Despite the fact that both Gedenet al. (1995) and the presentwork made use of housefly-derived isolates,no patent infections were induced on mature larvae living in microbe-rich substrates. Conversely,B. bassianaconidia mixed in manure-sawdustmedium causedhigh mortality of second-instarlarvae (Watson et al. 1995);younger house fly larvaeare likely more sensitiveto fungal pathogens,House fly larvaecan also be expectedto have effective inherentantimicrobial defenses since they inhabitmicrobe-rich decaying organic matter. Perhapsless likely, physicalremoval ofconidia from the cuticle during larval tunneling could causelow susceptibilityto P. fumosoroseusin this work. Pathogenicityof this P. fumosoroseusstrain to insectshas beenrepeatedly shown against greenhouse whitefly, Trialeurodesvaporariorum Westwood (Sanchez-Pefra andYinqtez-Jaime 1996), and P. barbatus (personalobservation). These results indicatepossible limitations to the use of P. fumosoroseasconidia as a housefly larvicide,even if high concentrationsof viable, pathogenicinoculum are directly deliveredto insects.Selection of highly virulent strains is a criticalneed. Dr. TadeuszJ, Poprawski(deceased, USDA-ARS) confirmedthe identity o/P. fumosoroseus.Dr, Bradley A. Mullens (Univ. of California, Riverside) kindly commentedon this manuscript.The statisticalsupport of Dr. Tom Bohman,Nate Martin, Pat Truxillo and ACITS (University of Texas at Austin) is greatly appreciated.This

224 researchwas supportedby Project 02-04-0902-2225,Direcci6n de Investigaci6n, UniversidadAutonoma Agraria Antonio Narro.

LITERATURECITED

Brownbridge,M., S. Costa, and S. T. Jaronski.2001. Effect of in vitro passageof Beauveriabassiana on virulence to Bemisiaargentifolii. J. Invertebr.Pathol. 77: 280- 283. Dodge,H. R. 1958. Domesticflies: pictorialkey to commonspecies in the U-S. U. S. Deparfinent of Health, Public Health Service, Communicable Disease Center TrainingBranch, Atlanta, Georgia. Geden,C. J., D. A. Rutz, and D. C. Steinlraus.1995. Virulence of differentisolates and formulations of Beauveria bassianafor house flies and the parasitoidMuscidifurax raptor. Biol. Control5: 615-621. Leucona,R. E., M. Caldeo,and D. C. Crespo.1996. Behavior of Beauveia bassianaand Metarhizium anisopliae on larvae and adults of Musca domestica,p' 48' In Proceedings 29th Meeting, Society of lnvertebrate Pathology, universidad de C6rdoba,C6rdoba, Spain. Majchrowicz, I., T. Poprawski,N. Maniania, and P. H. Robert. 1990. Effects of entomopathogenicand opportunisticfungr on Delia antiqua @iptera: Anthomyiidae) at low relativehumidity. Environ. Entomol. 19: I163-1167. Mullens, B. A., J. L. Rodriguez,and J. A. Meyer. 1987.An epizootiologicalstudy of Entomophthoramuscae in muscoidfly populationson southernCalifornia poultry facilities,with emphasison Muscadomestica. Hilgardia 55: 41 pp' Pereira,R., J. L. Stimac,and S. B. Alves. 1993.Soil antagonismaffecting the dose- responseof workers of the red importedfire ant, Solenopsisinvicta, to Beauveria bassianaconidia. J. Invertebr.Pathol. 6l: 156-161. SASSoftware, 1996. Release 6.12. SAS Institute Inc., Cary,N'C' Samson,R.A. 1974.Paecilomyces and some allied Hyphomycetes.studies in Mycology. 6.CBS, Baarn. l19pp. S6nchez-Pefra,S. R., and G. A. V6zquez-Jaime.1996. Mortalidad acumuladatras aplicaciones sucesivas de tres hongos entomopat6genossobre ninfas de mosca blanca, pp. 23-25./n Proceedingslgth CongresoNacional de Control Biol6gico, SociedadMexicana de ControlBiol6gico, Culiac6n, Sinaloa, Mdxico. Shields,M. s., A. J. Lingg and R. c. Heimsch.1981. Identification of a Penicillium urticae metabolitewhich inhibits Beauveriabassiana. J. Invertebr.Pathol. 38:374- 377. Steinkraus,D. C., C. J. Geden,D. A. Rutz, and J. P. Kramer. 1990.First report of the naturalocurrence of Beauveriabassiana in Muscadomestica. J. Med. Entomol.27: 309-312. Watson,D. W., C. J. Geden,S. J. Long, and D' A. Rutz. 1995.Efficacy of Beauveria bassiana for the control of house flies and stable flies (Diptera: Muscidae). Biol. Control5:405-41.

225 vol.28 NO.4 SOUTHWESTERNENTOMOLOGIST DEC.2003

AI.IALYSESOF COTTONLEAFCHARACTERISTICS EFFECT ON BF.\vIISIA TABACIGTOMOPTERA: ALEYRODIDAE) BIOTYPE B COLONZATIONON COTTONIN ARIZONAA}.ID CALIFORNTA

C. C. Chu, E. T. N*wickt, T.-Y. Che4 and T. J. Henneberry USDA, ARS, WestemCotton ResearchLaboratory 4135E, BroadwayRoad Phoenir AZ 85040-8803

ABSTRACT

Cotton leaf characteristicsand Benisia tofuci (Geffiadius) biotype B population density relationshipswere studied on 17 varieties of upland cotton, Gossltpiumhireilum (L.), in Arizona and California fiom 1999to 2002. Our objectiveswere to determinethe effects of trichome density, leaf are4 leaf perimeter, leaf perimeter-leafarea ratio, and okraJeafvs. normal-leafvarieties on populationsof B. tabaci in the field. The results showed that the density of branchedstellate trichomes on underleaf surfaceswas the primary factor influencing the varietal susceptibility to adult B. tabaci. Increased numbers of eggs and nymphs were found on hairy leaf Stonville (ST) 474 plants comparedwith smoothJeaf varieties. Leaf perimeter size and leaf perimeter-leafarea ratio were negatively related to B. tabac, numbers. Okra-leaf varieties had larger leaf perimeters and higher leaf perimeter-leaf area ratios and fewer adults per leaf and immatures (eggs and nymphs) per cm' of leaf disk. Smaller leavesof normal, smooth leaf varieties may result in lowerE. tabaci populationdensities.

INIRODUCTION

Bemisia nbaci (Gennadius)biotype B has been an economic pest in cotton, Gossltpiumhirwtum (L.), in the souttrwesternUnited Statessince l99l (Natwick et al. 1995). Significant lossescan occur becauseof reducedyields and lint contamination from honeydew excreted by adults and nymphs. A long-term solution to B. tabaci managementthat is environmentally and economically acceptable is needed. The developmentof B. tabaci resistantvarieties by conventionalplant breedingand selection are approachesthat warrant increasedattention for B. tabaci population suppression (Sippell et al. 1987). The adverseeffects ofresistant varietieson insect populationshave been dramaticwhen successfulhost plant resistancehas beenaccomplished (wiseman 1999). Our previousstudies identified cotton leaftrichome density,leafthicknesq and leaf shapeas potential traits that may be incorporatedinto upland cotton breeding programsfor the developmentof B. nbaci resistantvarieties (Chu a. al. 1998, 1999, 2000, 2001,2002). We reportherein the resultsofa four-yearstudy (1999-2002) on the impact of leaf perimeter sizes, leaf areas,and leaf perimeter-leafarea ratios as well as underleaf trichome densitieson colonization of B. tabaci on cotton varieties in Arizona andCalifornia.

^ Universityof CaliforniaDesert Research and Extension Center, Holtville, CA 92250. 235 MATERIALS AI.ID METHODS

All varieties, but G. tktrberi (Todaro), were upland cottons, G. hirsutum (Tfile l). Thoe were seven normal and ten okaJeaf varieties studied in the six'field experimentsfrom 1999to 2002. All normal-leafvarieties were developedin the United States. All okra-leaf varieties, except G. tfurberi, were developedin Australia. G' thrberi is a wild cotton found in at least eight countiesin Arizona and Northem Mexico 1960). All varieties were growing at 610 to 1,524 m elevations (Kearney and Peebles -The smoothJeaf cotlons except Stonwille (ST) 474 which is a hairy-lea{ yariety. smoothJeafvarieties averagedabout 20 or less stellatetrichomes per cmr leaf disk from fifth node main stem leaves-ascompared with l2O per cm' leaf disk for the hairy leaf ST 474 va;1ay. All varieties were not studiedeach yesr in each orperiment becauseof the "r,aitaUititi of seed(e.g., 93020-88-753,EO223,EOZS8, and Texas l2l)' The numberof varietiesstudied per year rangedfrom four to te{r.

TABLE l. Normal- and Olaa-Leaf Varieties S$died for SusceptibilityIo Bernisiatabrci B at Maricopa, AZ urd Ci\1999 -2002.

shape rype Oorlf SorH 1999-1 20fn,-2 2@l-3 DP 2OB N r DP 5OB N s x DP 5415 N s ; x DP 9OB N s x xx NuCOTN33B N S x xx sr 474 N H xx ; ; Torasl2l N s x E0223 o s xx x E0798 o S xx E1028 o S xx I G. tlurberi o S ; SiokraL23 o S x xx x x Siokra I-4 o s X S x FiberMax819 o : FiberMax832 o s x 89013-ll4 o s 93020-88-753-o s T x (( 3 Total N 5 22 o 5 44 224 ' N *d o d"*te okra- and normal-leafstrains or cultivars, respectively. o s *a H denotesmooth and hairyJeaf cultivar or strain,respectively. a X denotesthe varietiestested in experiments'

designs' The six experimentswere conducted in randomized complete block Arizona Agricultural Experiments l, 2, and I were condusted at the university of 4' 5' and 6 Rele.rctr Center, ivtoi"opt, Gzon4 from 1999 to 2001; and Experiments ExtensionCenter at ,"o. p.rfotr.A at ttre Uriiversit' of California Desert Researchand ExperimentsI and 2 Holwille, californi4 to* zooi'to-4. 2002. Someof the res'lts from set was analvzed for ;;;; ,"oon"J earlier (iiu "t. zoozl. The entire data perimeter,leaf area"and leaf' ;;.d;;r-;itn U*p",i."nt i to *tuav the input of leaf 236 perimeter-leaf area ratios and leaf trichome densities on B. tabrci. There were four replicates for Experiments l, 2, and 3; six for Expuiments 4 and 5; and five for Experiment6. Plots were eight rows wide for E:rperimentsl, 2, and 3; two rows wide for Experiment4; and four rows wide for Experiments5 and 6. Rows were 12.2-mlong and l-m apart with 3-m non-plantedbuffer areas between replicates in each experiment. Seedswere plantedand watered for gerrninationon 19 Agril 199, 13 April 2000, 16 April 2001, 2l March 2001, 22 Agnl 2002, and 22 April 2002 for Expoiments I to 6 respectively. Plarfs emeqgedabout trro urc€kslat€r and were wat€red at l0- to 20d intqvals during the growing sealr,ns.Plots w€re not tneatedwith any insecticidesorcept for diflubena[on (l{zLclrtorophenyl-3-(2,6difluorobeozoylurea))) for the codrol of salt marsh caterpillars,Esignerc rcrea (Dntry) (Lepidoptera:Arctiidae) on 13 August 1999 during E:periment l. Standardagronomic practices were followed for eachoperiment, Densitiesof B. nbrci on cotton were estirnatedat 7-d intervals from 2l July to 6 October in 1999 for Experiment 1, from l0 July to 5 Septemberin 2000 for Experiment 2, urd from 2 July to 20 Augrrst 2001 for Experiment 3. On each sampling date, three plants per plot were selectedat randomfor ExperimentsI and 2, and five plants per plot for E: 2.5-cm benreenthe two largestleaf lobes. The fifteenth node positions were at the lowest point on the main stemsat the time of sampling. A2-om2 leaf disk was taken from the area adjacentto the middle of the leaf primary vein at the baseofthe leaf(Naranjo and Flint 1994). Numbersofeggs and nymphswere countedon the undersidesof the leaf disks with the aid of a stereomicroscope.Adults per leaf were countedon the underleafsurfaces at the sameleaf positionson the main stemnodes using the leaf-turn method describedby Naranjo et al. (1995). Leaveswere randomly selected from plants throughout each plot. Leaf area and leaf perimeter of each sampled leaf during the growing seasonwere measuredwith a leaf area meter (CI400 CIAS Image Analysis, CID, Inc., Vancouver,WA). Densitiesof B. tabrci eggsand nymphson plantswere estimatedfiom samplesof ten plants selectedat random in eachplot bi-weekly from 5 July through 18 September 2001 for Expaiment 4, and from June through mid,Augrrst in 2002 for Experiments5 and 6. Leaf sampteswere picked fro-m the fifth plant node as described. Eggs and nymphs were counted on single l.Scm' leaf diskg as describedpreviously. Mults per leaf were countedon ten, fifth node leavesalso as describedprwiously. Seasonalmean numbers of B. tabqci eggs, nymphg and adults were analyzed using AI.IOVA When the treatment effects were significant, orthogonal comparisons were made between normal- and okraJeaf varieties with and without the hairyJeaf ST 474 vatiety data (Anonymous 1989). In additioq regressionanalyses were conducted comparingB. tabaci densitiesand trichome densities,leaf areas,leaf perimeters,and the leaf perimeter-leaf area ratios relationshipsusing stepwise procedure for the analyses (sAs 2000). Numbers of eggs and nymphs per cm' leaf disk were combined into immaturesfor the analyses.

RESI.]LTSAI{D DISCUSSION

Okra-leaf varieties had significantly fewer adults, eggs, and nymphs compared with normal-leaf varietiesin all paired comparisonsin the six experimentswhen the data for the hairy-leaf variety, ST 474, were included in the analyses(Table 2). However, this did not occur when the ST 474 data were excludedfrom analyses. tt appearsthat the ST 474 hairy leaf variety data had a disproportionateinfluence on the mean densities of

237 adults,eggs, and nymphson normalJeafvarieties. Butler et d. (1991)reported earlier that B. tabaci populations increased as the number of trichomes increasedup to 70 trichomesper 13.7mm".

TABLE 2. SeasonalMean t SE Numbers of Bemisia tabaci Biotype B Life Stageson All Normal- and Okra-Leaf Cottons at Maricopa, lg, 1999-2C/l.-1,Experiments l-3, and Holtville. CA. Experiments4-6, 2ool-2002. No No./cmz leaf disk Exoeriment Year Treatment adultVleaf Eggs Nymphs ST 474 included 1999 Okra 12.3+ 0.9bE 29.7r. L.gb 10.7+ 0.9b Normal 23.7*1.9t 37.0t 3.6a 15.3t l.8a 2000 Okra 3.6t 0.5b 6.0r 0.8b 2.4+ O.4b Normal 7.3* 0.8a 18.6t2.6a 6.1t 0.9a 2001 Okra 7.6tl.2b 6.9+ l.ob 1.5* 0.2b Normal 15.7t23t 18.l it4.2a 4.4*l.Oa 2001 Oka 1.9i 0.6b l.l + 0.3b 0.6t 0.2b Normal 7.6t0.1a 10.5t 2.0a 6.0tl.2a 2002 Okra 0.3+ 0.lb 0.2r 0.6b 0.210.5b Normal 2.8i 0.5a 2.8t 0.6a 2.9t 0.5a 2002 Okra 1.8r l.0b 0.5r 0.4b 1.2t 0.5b Normal 7.1t l.2a 2.3!.0.4a 3.6t 0.6a ST 474 not included I 1999 Okra 12.3i 0.9b 29.7+ l.9t 10.7* 0.9a Normal 2l.Otlja 31.3+2.6a 12.0t l.0a 2 2OOO Okra 3.6 i 0.5b 6.0r 0.8b 2.4t0.4b Normal 5.810.3a 13.6t l.Oa 4.4t0.4a 3 2001 Okra 7.6t l.2b 6.9r l.ob 1.5+ 0.2b Normal ll.l t 1.0a 9.1t l.0a 2.3XO.2o 4 20Ol Okra 1.9t 0.6a 1.1r 0.3b 0.6+ 0.2b Normal 5.6t 0.5a 4.4+ O.3a 2.3tO.2a 5 2OO2 Okra 0.3t O.lb 0.2+ 0.la 0.2i 0.la Normal 1.2t 0'1a 0.4+ 0.la 0.7t 0.la 6 2OO2 Okra 1.8t 0.2a 0.5t 0.la 1.2t 0.lb Normal 2.7XQ.3t 1.0t 0.la 1.4t 0.6a son in an experiment not followed by the same letters are significantlydifferent by orthogonalcomparison (P: 0'05)' For ST 474 included,thJF valueswere from ,i.s a a$ with P < 0.01 for ExperimentsI to 6 for adufts,eggs, and nymphs,respectively. For ST 474 not included,the F valueswere 43.9, 0.2, andv3'wrth p < b.ol, p > 0.05,and p > 0.05for Experimentl;22.9,38.9, and20.1 < witir p < O.Ol,P < O.Ol,and P < 0.01 for Experiment2; 16'1,5'8,rnd 22'8 vtlthP 0.01,P<0.05:andP0'91'P<0'05, > and'P<0.05 fjorExperiment 4; 5.2,'0'6,and 0.8 with P < O'05,P > 0'05,and P 0'05for Experiment5;2.7,0.t, and5.9 with P > 0.05,P > 0.05,and P < 0'05 for Experiment6' for adults, eggs,and nymphs,respectively.

Regressionanalyses suggested that when all factorswere considered,the equation accountedfor 66.1 and 68.0% of the variation in numbersof adults per leaf and ir.",r*t per cm' of leaf dislq respectively(Table 3). Underleaftrichome density was in the most important factor measuredand accountedfor 53.2 and 62.2Yoof the variation 238 IAILE 3. RegressionAnalyses of TrichomeDensity, Leaf Area" Leaf perimeteq Leaf Perimeter-Leaf and Area Ratio of CottonLeaves and the Density of Bemisia tabaci B life 1999-2002. Leaf Leaf perimeter- Trichome leafarea ratio" Total R2 Aduhdleaf 19.36 + 0.107 Partial I 0.532 ImmatureVcm2leaf disk 38.426 + 0.318 - 0.379 Partial I 0.622 0.058 0.680b " Leaf perimeter and l - b["fl:1:-l:!:'aI9" and immature/crn2teaf dish ,"rp""ti"r"iv. benotessig*ncant |r.{uttdteafatF= o.OOt

nymlers of adults per leaf and immature per crn2leaf dish respectively. other authors {so lave reportedthat cotton varietieswitti high trichome density "." .o." -rssr,zusceptible to B. tabaci colonization comparedwith smooth-leafvarieties @uher J "1. Norman -andSparks 1997, Chu et al. 2000). High trichome density is associatedwith increased boyndqv layer humidity on leaf surfaces@unage rezr;, wtrictrmay u"l,nportant ro, r. tabaci indesert condition habiats. Leaf perimeter-leafarea ratio acc,ounteifor l2.g/o of the variation in numbersof B. abaci aduitsper leaf Large leaf perimeter-leafarea ratios areusually associated with smallerleafareas that allow l"tto # circulatioiin the plant canopy and possiblyless boundarytayer humidity under desertconditions and tlius a more unfavorablehabitat for the 8. tabaci adults. iegression analysesalso indicatedthat leaf perimeter was significantly(5.8o/o,F:19.6 wlttr F < 0.001)reiated to the variationin numbersof B. tabaci immatures(Table 2). Theserezults reemphasizethe importanceof the smooth leaf cotton characterfor reducingS. tabaci cnlonization. OkraJeafcotton has also been suggestedas a sourceof resistanceto B. tabaci and twospotted spider mite colonization 6filron 1994, Jenkins 1999,chu et t1..2o02). Sincealmost all uptandcottons planted in the united Statesare normal-leaf varieties, when B. tabaci is an economic pisg decreasingleai size of the normal-leaf cotton to increaseleaf perimetersand leaf perimeter-areaitios might be an additional effective strategyto reduceg. taboci colonlzation.

ACKNOWLEDGMENT

We thank Hollis M. Flint and C. Glen Jacksonfor their constructivereviews of an earlierversion ofthe manu_scriptAppreciation is alsoextended to patrick J. Alexander, Eleanor R. Gladding, and Scott DaviJ for their technical assistancewith the study. The seed.swgre_ by: ryordeq cotton SeedDisfiibuton Ltd., Dalby, eueensrand,A'sharia; plbgt_d (orka-leaD variety including the three E's by austraiia's Commonweatttr Scientific and Industrial. Research organizatioq Narrabri, Australia; FiberMax by Aventis lolmerlv cropscience,Research Triangle park, NciDp anaNucorr.r 33B by Delta and Pine Land cg., scot! MS; Texai l2r by South Texas pranting Seei Associatio4 Mercedes,fi; andG. thurberiby university otcariromia, nivlrside, cA.

LITERATI.JRESCITED Anonymous. 1989. MsrAT-c. A microcomputerprogram for the design,managemenq and analysis of agronomic researchexperiments. Michigan state-univenit-y, rasi Lansing, ML

239 Burragg S. W. 1971. The microclimate8t th€ leaf surface,pp. 9l-101. .ln-T.F. Preece -a c. n Dickinson [eds.] Biology of leaf slrfrce organisms. Academic,London. Butler, G. D., Jr., F. D. Wilsoq and G. Fisher. 1991. Cotton leaf trichomesand l0: 461464.. population'c.-c., s of Enprca tybica ud Benisio.tubrci. cropPrd. Henneberry. 1999. chu,' a c. coheq E. T- Natwich G. s. simmons,and T. J. Bemisia tabaci (Ilqnipaera: Aleyrodidae) biotype B colonization and leaf morphologyrelationships in uplandcotton. Aust.J. Entomol. 38: 127-131. P' walker' chu''and c. i', r' i. Freemaq r' s' nuitner, T' J' Henneberr'' D R Nelson5G' b. T. N*wick. 2OOO. Bemisia wgentitolii Qlomoptera: Aleyrodidae) colonization on upland cottons and relationshipsto leaf morphology and leaf age. Ann. Entomol.Soc. Am. 93:912-919- Chu, C. C., T. P. Freeman,J. S. Buckner,T' J. Henneberry,D. R.Nelson, and E' T' Natwick. 2OOl. Susceptibility of upland cotton cultivars to Bemisia tabaci (tlomopterra:Aleyrodidae) biotype B in relation to leaf age and trichome density. Ann. Entomol.Soc. Am. 94:743'749. Chu,' C. C., E. T. Natnick and T. J. Henneberry. 2002. Bemisia laDaci Qlomoptera: Aleyrodidae) biotype B colonization on okra- and normal-leafuplandcotton strains andcultivars. J. Econ.Entomol. 95:733'738. J' Chu,' C. C., E. T. N*wicb H. H. Perkinq D. E. Brushwood,T. J. Henneberry,S' castle, A. c. cohen, and M. A. Boykin. 1998. Upland cotton susceptibilityto Bemisia ogmtifotii (tlomoptaa: Aleyrodidae)infestations. J. Cotton Sci. 2: l-9. Jenkins,l. N. fSSS. Host planiresistancein cotton and its value,pp. 45-58. InI. ^. wisemand and B. R. webst€,f [eds.] Proc. ThomasSay Publications in Entomology, Entomol.Soc. Am., Lanham,MD. Kearney,T. H., andR. H. Peebles.1960. ArizonaFlora. Univ. california Press,Berkeley andLos Angeles,1085 PP. Naranjo, S. E., anil H. M. nda. 1994. Spatialdistribution of preimaginalBemisia tabaci sequential fttoropio", Aleyrodidae)in cotton and developmentof fixed-preoisio4 sampling'S. plans, Environ.Entomol. 23: 254-246. of adult Naranjo,-Bemisia n,'H. M. Flint, and T. J. Henneberry. 1995. Spatial distribution tabaci in cotton and development and validation of fixed-precision sequentiatsampting plans for estimating population density. Environ. Entomol- 24:261-27O. Natwick, E. T., c. c. chu, and T. J. Henneberry. 1995. Pima and upland cotton susceptibiiity to Bemisis ogentiftlii und€r desertconditions. Southwest.Entomol. 20:429438. Norman,J. W., andA. N. Sparks. 1997. Cottonleaf hairsand silve,rleaf whiteflies' in the lower Rio GrandeValiey of Texas. Three-yearresearch summary, pp. 1063-1064. 1z R. Dugger and D. A Richter [eds.] Proc. Beltrn'ideCotton Conf., Natl. Cotton Councilof Amer.,MemPhis, TN. SAS Institute. 2000. SAS/STAT user's guide. SAS Institutg Cary, NC' g. Resistanceto whitefly (Bemisia sippell," D. w., o. S. Bindrq ana xtralita. 1987. t6baci) in cotton (Gossltpiun hirvtum) in the Sudan. Crop Prot. 6: l7l-178. Wilson, L. t. lgg4. Resistanceof okraleaf cotton genotype.sto twospottedspider mites (Acari: Tetranychida€).J. Econ'Entomol. 87:172o--1735' wisemaq B. R. lbgg. 5u""".. in plant resistanceto insects,pp. 3-15. In l. A. proc. Wisernandand B. R. Webster teds.l ThomasSay Publications in Entomology, Entomol.Soc. Am., Lanhaq MD.

hrdlased bYthe U.S.DeparfiEnt of Agtiq.rlbre FOR,OFFIGAT I.|SE OI{LY

240 vol.28 NO.4 SOUTHWESTERNENTOMOLOGIST DEC.2003

AERIAL MOVEMENTS OF THE COTTONAPHID, APHIS GO^'STPI/GLOVER (HOMOPTERA: APHIDIDAE),AND DISCOVERYoF MALE MORPHSIN THE TEXAS HIGH PLAINS

M. N. Parajulee,tp. W. Kidd,tD. R. Rummel,randW. p. Morrison2

ABSTRACT

Aerial dispersalof the cottonaphid, Aphis gossypii Glover, was monitored with an Allison-Pike@insect suction trap at the TexasAgricultural Experiment Station, Lubbock, Texasin 1989,1992,1993 and 1994. This suctiontrap collecteda single maleA. gossypii on 2 December 1992, lhe first confirmed report of the male form oi l. gorrypit in-ihe TexasHigh Plains. A combinedtotal of 13 male morphswere collectedly ihe suction trap and from cotton plants near the trap site. In an effort to establish a method of predictinginfestations in cotton,cotton aphid suctiontrap collectionswere comparedto field densitiesin a nearbycotton field. Resultsindicated that suctiontrap coilections duringthe growingseason were not predictiveofcotton aphid field populationdensities. However,numbers of cotton aphids collectedby suction trap in Novembermay be predictiveof springfield populationdensities.

INTRODUCTION

The cotton aphid',Aphis gossypii Glover, was first describedby TownsendGlover in 1876(Palmer 1952). Two commonnames (i.e., cotton aphid and melon aphid) are currentlyaccepted for A. gossypii. A. gossypiiis a widely studiedpest of interesrro many producersand researchers.lnfestations by the aphid in cotton may be consideredan intermittentproblem; in someyears the aphidis a secondarypest (Leser1994), while in other years,years in which widespreadoutbreaks occur, the aphid is a severeand economicallysignificant pest (Leser et al. 1992). Currently,there are no accuratemethods of predictingthe densityto which aphidpopulations will developin cottonfields. The life cycleof the cottonaphid in the TexasHigh plains,as well as acrossmost of the insect'srange, is complexand not fully understood(Kring 1972). Most of themany variationsin the life cycle of l. gossypiiare known to occur in the TexasHigh plains, whereit is thoughtto consistprimarily of viviparous,parthenogenetic females (-slosser et al. 1989). A holocyclicstage has been recorded in Connecticutwhere cotton aphids used Catalpabignonioides walter (commoncatalpa) and Hibiscus syriacus L. (rosetrsuron; as primary hosts (Kring 1959). o'Brien et al. (1990)collected four oviparousand two male cotton aphids from cotton regrowth at the Delta Branch Experiment Station in Stoneville,Mississippi, on 20 November1988. Theseauthors also reported a total of six malesand eight oviparousfemales in the National Collectionof Insectsat Beltsville. Maryland.The possibilityof a holocyclicstage has not beenexamined in theHish plains.

I Texas Agricultural Experiment Station, Rt. 3, Box 219, Lubbock, Texas 79403 2 Texas Cooperative Extension, 2475 T ANllJ, College Station, Texas 77843 241 The distributionof the cotton aphid is worldwide,mainly due to its broadhost range(Blackman and Eastop 1984). Cottonaphids are highly pollphagous,utilizing plant speciesbelonging to 89 families(Ebert and Carhvright 1997). Theyare known to transmit over 50 plant viruses(Kranz et al. 1977). Otsrien et al. (1992)collected cotton aphids from 24 non-cultivatedhost plants representing18 plant families in Mississippi. The majority of thesehost plants occurredwhile cotton was present. Cotton aphidswere collectedfrom at least two of thesehost plants during every month of the year and overwinteredon three species: eveningprimrose, Oenothera speciosa Nuttall; henbit, Lamiumamplexicaule L.; and dock,Rumex spp. A surveyof cottonaphid host plants has not beenperformed in the Texas High Plains,but many of the documentedhosts are commonin this area. Cotton aphidsappear to exhibit a form of aestivationin the TexasHigh Plains. Small,yellow forms are observedunder unfavorable (hot and dry) conditionsduring the sunmer. Theseforms may not grow or reproduceuntil conditionsbecome more favorable (Kring 1959,Kranz et al. 1977). Both winged and winglessforms are commonin the TexasHigh Plains. The exactcause of productionof alateforms is not fully understood, but it seemsto be primarily associatedwith daylength(Slosser et al. 1989). Loxdaleet al. (1993) statedthat productionof alatesis "an environmentally-inducedexpression of genotype"and that not all wingedforms actually disperse. Althoughcotton aphid outbreaks in the TexasHigh Plainsare believed to be aided by large influxes of aphids from more southerly locations, this has never been documented.Long-range dispersal ofcotton aphidsis poorly understood,perhaps due to their smallbody size. Aphid flights over greatdistances may not occurby choice. Aphids may developlarge populationsin localeswhere suitablehost plants are availableand subsequentlybe carriedby prevailingwinds (Loxdaleet al. 1993). The exactimpact of theseimmigrants on TexasHigh Plainscotton aphid field populationsis unknown. There is a paucity of informationon the life cycle of the cotton aphid and the impactof influxesof aphidsfrom moresoutherly locations in the TexasHigh Plains. The objectivesof this study were to investigatethe aerial movementof the cotton aphid throughoutthe year,to relatethese findings to field populationdensities, and to examine thepossibility of a holocycliclife stagein theTexas High Plains.

MATERIALS AND METHODS

Aerial cotton aphid populations were monitored at the Texas Agricultural ExperimentStation, Lubbock, Texas, by using an Allison-Pike@insect suction trap (Ailison and Pike 1988)in 1989,1992, 1993 and 1994. Specimenswere captured in pint jars containinga mixtureof ethyleneglycol (10%)and either ethanol or water(90%). The irup ** conitructed from a 1.5-m sectionof 38-cm polyvinyl chloride (PVC) pipe connectedto a 6-m sectionof30-cm PVC pipe usinga coupler. Suctionwas created by a 30-cmelectric fan poweredby a 38-wattelectric motor mounted under the dollectioncone andjar in the lower portion of the trap. The trap sampledapproximately 570mr of air per hour from 8.3m above the soil surface. Surveyswere conductedweekly from June throughoctober in 1989,April throughDecember in 1992and 1993,and from March though Decemberin 1994. All aphidswere extracted from thesesamples, counted, and storedin 70% ethanolfor identifiiation. Cotton aphidswere then separatedfrom other species.voucher specimenswere placed at theTexas Tech University Museum. cotton aphid seasonalaveiage data (numberof aphidsper leaf per week) from untreated,inigated cottonwere obtainedfrom experimentalplots adjacentto the suction trap site in t-gSZ-tgqSin an effort to establishthe relationshipbetween cotton aphid abundancein the field andnumbers ofcotton aphidscollected bythe suctiontrap' Cotton

242 on leaves aphidswere sampledweekly in the field by countingaphids ihree.fufllsized io*e. third of eactr-of ten plants (total 30 leaves). picked'Su-pting from the upper,-iJfu", *a ** conductedfrom 15 July to 25 Augusteach year' RESULTSAND DISCUSSION timing of occurrenceof Suctiontrap collectionsof cottonaphids showed that the ofthe aerialactivity in the first cottonaphid, seasonal activity pattems, and the cessation peakflight activityoccuned the TexasHigh plains "-l.J *ong yeis. Nevertheless,the e-!)' Peakactivifv periods in mid- or late August in uif io* f"urs of the study(Fie' I *.* g"""tdfy of sf,ort durations,lasting 1-2 weeks,except in1992' the_lastcotton aphids The first cottonaphid coilectedln 1989was on 8 July, and p""t during the week of were collectedon f z s"pi"mU.i. flight activity in ry.8999:utted 1f)' 2i Augustwitfr 1,982cotton aphids captured tl.tfre trap@ig' this was the only kt 1992,tne frrst cottin aphid was collectedtn 13 May, and Flight activityincreased in late ,peci-en *tt""t.o in rgsi be;een 15 April and2l July. with 305 aphids-captured per trap i"riv *J,rr" peak activitywas recordedon 12 August of Augustat the per week (Fig. lB). cotion aphidswere captured throughout the month flight activitv in 1992 I"t. of "ppr.-*imaiely foo "prJiJr per trap.plr week. Cotton aphid captured-onthat date' ..*.a "i, Z Decemberwith irr. 1,1"t",p..itt * of thecotton aphid Thefirstcott"ctionoracottonaphioinlgg3occunedon30June.Numbersthen per on.25 August(Fig' lC)' increasedsporadically to a peak of 40 aphidsper trap week reaching24 per y,apon 77 Cottonaphid numbers continued to vary from weekio week, Novernberand samplingwzrs s"pt..u".. The last cotton aphid was collectedon 25 discontinuedon 21 December. No othercotton kr lgg4,the first collectionof a cottonaphid occuned on 26 May. began to increase,and aphids were collected *iii ii July (Fig. lDj. Numbers then the following pi"i.a "i O:O aphidsin the hap on )S e,ugus; Flight activity declined gradually to one aphid on 24 week, and the aphid numberscaptured by thi traq declined September'onlysevencottorr'aphids'o.coll".tedbetween24Septemberand28 and2 December'During O.ioU.., but a total of 321 werecollected between 6 November (Fig. 2). No cotton this period,the higlrestweekly count of 131occuned on 17November and the last samplewas "pfrfi, *rir collJctedby the suctiontrap after 2 December, collectedon 29 December. the four yearsof The earliestcollection of a cotton aphidby the suctiontrap within of. a cotton aphid in the study wns on n M;;-l;gr. rrt" aut. of tire first appearance populations' Peakcotton ;;t ; tap samplesdid not appearto be predictiveof field the secondor third week of upilio .oil"rtlo"s'by the suctiontrap typically occurredduring late July.to late August' iugu.t. The datis of peak field populationsranged from untreatedfield population orJrf"ppl"g the periodoftrigh suctiontrap counts.The highest rherefore' ;;;'t6;;: recordedin Lsis (Fie. 1C), when tra.pcollections Y=l::.1tjfield infestations' tr"p .ott..tiorrs duringthis perioddo not appeaftobe predictiveof appearedto The increasel" .ottin aphid collectionsby thJ suctiontap in late July locations and representan increasingimmigrltion of cotton.aphids from more southerly it is also increasedaerial movementoT .ottott aphidsin the Lubbock area' However, Lubbock area that the locatized aphid infestationsin pre-squaringcotton in the f"$itr" theseaphids could might haveproduced .o*. ulut. forms in late Juneand early July and tranin mid- alsohave been caught on suctiontraps. lncreasesin aphidcollections !1 Jhe to late August and -arly Septemberappeared to representlocal cotton aphid dispersal. A-second peak in'cotton upttia collectionswas noted in late october to mid- aphids out of mature NovemberGig. Zi. This peak appiared to representmovement of

243 -.- -+ SuctionTrap FieldAbundance

l0

&0() 0g q) o) B ;,rr q) r- 30 B q) 50L 4) 409 !! 200 G ti F 30 .9 loo 20(D I r04 0 q)10 oa o ,10 150 L E q) 430 R 100 = Zzo z L o) o0 =10 s0 !l H c) '4.= 0

.aEE==s5$$883 tt) =====ts g 3,g g r€ E 3 F; i fi r N SampleDate FIG. l' Numbersof cotton aphidscollected per week by a suctiontrap and mean numbersof aphidsper leaf in untreatedcotton plots, Lubbock,Texas, 1989, 1992- 1994. -.- SuctionTrap + FieldAbundance

,L 6) (D 50 ,xq) F 6) L I F 0EL L F 12 ii

L 9 eE (t) 0 ()L 6-=

v,

L oR

L q) 4sz tr

z 30

II**G:ERA8s88tEEEA FEig$i$eFK,At9*

Sample Date

fall FIG. 2. Numbers of cotton aphids collected per week by a suction trap during.the the unJ rn.un nu-bers of coitoh aphids per leaf in unUdated cotton plots during following surrmer. Lubbock, Texas, 1992-1995.

245 cotton,which was no longernutritionally suitable. Thesecotton aphids were believed to be searchingfor a suitableoverwinterins host. The first male cottonaphid recoided in the TexasHigh plainswas collectedby the suctiontrap on 2 December 1992. A total of sevenmale cottonaphids were capfuredin the suction trap over the four year_study period (Table l). All matesfound in suctiontrap sampleswere collected in Novemberand early December. In addition,a total of six male cotton aphidswere collectedin Novemberof 1:9sqfrom cotton prutttr *fti"rt had been grownin a caged environment.Identification of theseaphids was connr*.J uy Dr. S. E. Halbert (Florida Department of Agriculture,Gainesvilie, Florida). itre proouctionor malesin the fall impliesthe existenciof a sexuallyreproductive overwintering generation. However, no oviparaewere found on cottonor in suctiontrap sampres.;;;. specimen gossypii 9!ya\e,a, wascollected by a suctiontrap near parma, Idaho, on l2 5ctober l9g7 (S.E. Halbert, unpublisheddata). other confirrnedfindings or-ute'.otton iphids in the UnitedStates areshown in Table 1. All of thesemales weie collectedbetween 9 October 20 November, gd and collection sites were widely distributedthroughoui tt " urrir"a States. Therefore,production of malesin Novemberand early D"cemier appearsto be relatedto shorterdaylength. _ The purpose of male cotton aphid production is believed to be for sexual pqrodugtiol with oviparous females. Arthough no oviparaewere found in this study, a holocyclic life cycle likely exists in the TexasHigh plains. Sexualreprodultion would allow geneticrecombination to occur and increasethe size of the genepool. Further studies w^ouldbe requiredto identiff possibleoverwintering host plaits andexamine the impactof overwinteringaphids on field populations.

TABLE l. Known collectionsof MaleAphis gossypiiin the united Statesto Date.

SurveySite SurveyDate Host Plant Male Specimen Greenbush,MA 18Oct. 1949 strawberry I Washington, DCI 09 Nov. 1962 roseofSharon 1 Washington,DC' l0 Nov.1964 roseofSharon 2 Albuquerque,NM' 09 Oct.1969 catalpaleaves 2 Parma,ID' l2 Oct.1987 suctiontrap I Stoneville,MS' 20Nov.1988 cotton (regrowth) 2 Lubbock,TX 02Derc.1992 suctiontrap 1 Lubbock,TX 08 Nov. 1994 cotton 6 Lubbock,TX 17Nov.1994 suctron trap Lubbock,TX 02Dec.1994 suction trap 2 O'Brienet al. (1990). -S. E. Halbert,unpublished data.

Ifoverwintering -_ cotton aphidsinfluence early seasonfield populations,aphids collectedby a suctiontrap in Novembermay havea predictivevalue. Trap collections were terminatedearly in 1989, so it is not possibleto determinethe impict on 1990 infestations.The Novembertrap collectionsin 1992and 1993were relativelylow, and the 1993and 1994field infestationswere light, exceptfor 16 August in l99j @ig. 24, 2B). The Novembertrap numbersin 1994were approximatelythree-fold those of the previoustwo years, and unusuallyhigh cotton aphid field populationdensities were presentin 1995;the field infestationin 1995remained high throughoutAugust (Fig.2c). Thesedata indicatethat there exists a relationshipbetween the Novembelsuction trap

246 catches,the potential overwintering population, and the level of field infestation the following summerin the TexasHigh Plains. However,this informationis inconclusiveat this time due to limited amount of data. The year-to-yearvariation in abundanceof potentialoverwintering host plants,perhaps a function of late seasonrainfall, may.be moreimportant than numbers of late seasonflying aphids,as caught in the suctiontrap. In addition,winter severitymay also impact winter survival of cotton aphids. Additional studies are needed to capture year-to-yeaxvariation in the relationship between overwinteringpopulations and the resultingpopulation dynamics the following growing season.

ACKNOWLEDGMENT

Drs. Larry Sandvol and Susan Halbert (University of Idaho R&E Center, Aberdeen,Idaho) provided facilities and assistancefor identificationof specimens.We appreciatethe reviews of the earlierversion of the manuscriptby StanleyCarroll and Mark Arnold (TexasAgricultural ExperimentStation, Lubbock, Texas). This research wasconducted in partialfulfillment of requirementsfor the M.S. degreefor P. W. Kidd at TexasTech University, Lubbock, Texas.

LITERATL]RE CITED

Allison, D., and K. S. Pike. 1988. An inexpensivesuction trap and its use in aphid monitoringnetwork. J. Agric.Entomol' 5:103-107. Blackman,R. L., andV. F. Eastop. 1984. Aphidson theworld's crops: An identification guide. JohnWiley andSons, New York, NY. Ebert,i. A., and B. Carhvright.1997. Biology and ecologyof Aphis gossypiiGlovet (Homoptera:Aphididae). Southwest. Entomol. 22:1 16'1 53. Kranz, J., H. Schmutterer,and W. Koch. 1977. Diseases,pests and weedsin hopical crops. JohnWiley andSons, New York, NY. Kring, J. S. tssq. The life cycleof themelon aphid, Aphis gossypil Glover,an example of facultativemigration. Ann. Entomol.Soc. Amer. 52:284-286. Kring, J. B. 1972. Flight behaviorof aphids.Annu. Rev.Entomol. 17:461'492' Lesei l. f. 1_994.Management of cotton aphids: Texas style, pp. 137-l4l' In Proceedings,Beltwide cotton conferences,Nat. Cotton Council of Amer., Memphis,TN. L,eser,J. F., C. T. Allen, andT. w. Fuchs. 1992. Cottonaphid infestations in westTexas: A growingmanagement problem, pp. 823-827..In Proceedings,Beltwide Cotton Conferences,Nat. CottonCouncil of Amer.,Memphis, TN. Loxdale,H. D., J. Hardie,S. Halbert,R. Foottit,N.A.C. Kidd, and C. I. Carter. 1993' The relative importanceof short- and long- rangemovement of flying aphids' Biol.Rev.68:291-311. O'Brien, P. J., M. B. Stoetzel,and D. D. Hardee. 1990. Verifrcation of male and oviparous morphs of the cotton aphid in mid-south cofton (Gossypiumhirsutum L.). J. Entomol.Sci.25:73-74. O'Brien,P. J., M. B. Stoetzel,B. R. Leonard,and J. B. Graves. 1992. Taxonomyand biology of Aphis gossypiiGlover in the mid-south,pp. 646-648.,In Proceedings, BeltwideCotton Conferences, Nat. CottonCouncil of Amer.,Memphis' TN. Palmer,M. A. 1952. Aphids of the Rocky Mountainregion. A. B. HirschfeldPress, Denver,CO. Slosser,J. E., W. E. Pinchak,and D. R. Rummel. 1989. A reviewof known andpotential factors affecting the population dynamics of the cotton aphid. Southwest. Entomol. 14:302-313.

247 vol.28 NO.4 SOUTHWESTERNENTOMOLOGIST DEC.2003

LADY BEETLES CONTROL OF GREEN PEACH APHIDS IltyZUS PERSICAE' WftH HARMONIAAXYNDIS ON CHILE CAPSICUMANNUrt IN THE GREENHOUSE

Southwarda D.R. LaRock3,Zohair Mirdad,, J.J. Ellington" Traceyca:rillo3, andlvlorris Departmentof Entomology,Plant Pathology, and Weed Science New Mexico StateUniversitY Las Cruces,New Mexico 88003-8003

ABSTRACT

ladybeetles' chile plants,capsicum annuz L., in greenhousecages were inoculatedwith of I :20,l:40, Harmoniaaxyridls (Pallas), to greenpeach aphid, Myzus persicae (stizer),ratios eachplant in each 1:g0,1 :160, i :320,and t :6+0, rive aphidswere counted every two dayson six treatnentsatp cage'forl0 days. Changesin aphiddensitywere statistically siggificant in all M' J4.05.Treatnentratios of 1:j0, l:40, 1:80,1:160, and l:320gave9'o/oorbettercontrolof persicae afterI 0 days. The meanpercentage aphid reduction at a 1:640 ratio fell to 85.69%after period' tO daysUut still prwided sigrrificantaphid density dedeases during the 10 daytest

INTRODUCTION years chile, capsicumannuml. (Solanaceae),has been grown in New_Mexicofor 400 (Bosland et at. if1. A variety of pests damagechile both in the field and greenhouse; iowever, the greenpeach aphid Myzus perstcae (Sulzer) is oneof themost important' Mackauer pestin U. S' andWay (197e reiorteOftrat ttt" greJnpeach aphid is the mostcommon aphid g."enhoos".. fitis upniA also veJtors many plant viruses (Kennedy et al. 1962, Hanis and il{aramorosch 1977, Matthews l99l) and ii resistant to a large number of insecticides costofchemical (Georghiou1986, Quaclia et al. 1993).Due to pesticideresistance, escalating "ont o-1,and environmentalconcerns, interest in biocontrol strategiesis growing. Biologicalcontrol can be definedas the use ofone organismto decreasethe population all of anotheroriarrism (Huffaker and Messenger1976). It is often not n@essary,toeliminate pests,but onf to reducetheir numbersbelow an economicthreshold (Dent 1993)' Predatorsare iistri-tuteOamong 20 insectorders and are categorized by taxonomicaffiliation (Hodek 1973' Canardet al. 1984,Gilbert 1993,) or feedingstrategy, monophagyversus pollphagy(Hagen et al. re76). Lady beetlesare one ofthe most formidableaphid predators (Hagan and van denBosch 196g,Hodei 1973).The coccinellid.F/a rmonia aryridis (Pallas)is anintroduced species from theFar East(Timberlake 1943, Chapin and Brou 1991)and is establishedin thesouthem U' S.

'Homoptera: Aphididae 2Coleoptera : Coccinellidae 3Departnent of Entomolog5r,Plant Pathology and Weed Science,New Mexico State University' Las Cruces,NM 88003. aAgriculngal Biometric Service, New Mexico State University, Las Cruces, New Mexico. 249 a{ !!-t9nin erou l99l).. This speciesis highly polymorphicand may live up to threeyears (Hodek1973, Knodel and Hoebeke 1996). kt this study, I/. aryridis was evaluatedas a biocontrol agent for M. persicae in the greenhouse'The research sp_ecificallyinvestigated the optimal preiatorlp.ey ratio requiredfor aphidconhol within a specifiedperiod of timi,

MATERIALS AND METHODS

This research wasconducted at theFabian Garcia Research Center, New Mexico State U,niversity. Adult F/. aryridis were imported from Georgiato establisha laboratorycolony. Youngprogeny from thiscolony were used to carryout theixperiments. A laboratorycolony of H' persicaewas propagated from local insects,and was rearedon pottedchile plantsin two greenhousecages. "Nu Seedsof c. !!1num Mex Big Jim" weregrown'in 2.5-cm preformed peat potsuntilthethree-leafstage.Thehealthiestseedlingsweretransplantedintol2-cmplasticpots containingpeat mossand pearlite. All plantswere fertilizedwith a completewater soluble fertitizere0-20-20) andinigated oaily. err prantswere sprayed w**r:1T:c$:p,h:.plorus-porassium r,ayreton' tungicide Il-(4-chlorophenoxy)-3-3-dimethyl-l-(lH-12,4 nazot-i_g1_z butazone @ayercorp. Kansascity, Mo.)l at a-rateof1.52 mi activeingredient per 3.7gL of wateronce everytwoweeks to preventthe growth ofpowderymlldew,Leieiilula tauricaQ-nv.), Plantswere grown undera 15-hour light:9-hour dark photoperiod,After threemonths (March to June)'the plants averaged36.5-cm in height.The number of lruuesp". plantranged from 6g to 41-cm,averaging 53.5-cm. Eachexperimental unit consistedof four individuallypotted chile plantsenclosed in a screenedcage' Beforethe startofthe experiment,each pot ofchile wascovered at thebase of the plant with a small meshnylon net to preventinsects from falling onto the soil of the pot. Actualnumber of H. aryridis to M. persicaeper treatment were: (l) 4:g0,(2) 4:160,(3) 4:6a0, (5) a:1280 (6) and 4:2560.The experiment was run asa completelyrandomizeddesign. Each tr€ahnent was replicatedthree times (threecages containing four individually pottedchile plants). Controlswere established for eachtreaftnent (trvo cages containing fourindividually potted chile plants,but no H. axyridis). Cagesfor treatmentsand controlswere randomly assigned. The mean numberof M. persicaesurviving the predationof H. axyridis were determinedevery 2 daysfor a rO-dayperiod. At the enJof the l0-dayperiod, alrM. persicae were removed from eachcage and the cageswere inoculatedwittr ttre inaicatednumbers of predatorsand hostsfor the next treatment. Mean M. persicae densitiesand Standarderors were calculatedfor the treatmentsand controls. A dummyvariable regression @raper and Smith 1966)was usedto evaluatethe effectsof the variablestime, replication,and iime uy replicationinteractions. This analysis provides estimatesfor theparameters representea by itrevariaUles. The analysisestimated the intercepts andslopes for eachtreatrnent and furtherestimated the difference between the slope of the and .9o-nfols the slopeof the treahnents. Only five cagescontaining chile plantswere availablefor this experiment, thusthe treahnents weie run sequentially.Sigf,ificant -on"-*uydifferences were determined by combiningtreatment days 2, 4,6, g and l0 usingu analysisof variancemultiple range test.

RESULTSAND DISCUSSION

Table presents 1 themean density of M. persicaeper plant in thethree replication cages and two control cagesof eachtreatment as well asthe meanpercentage aphid riduction for all six ratio treatments.

250 TABLB l. MeanNrmber of M. persicae(t SE")Counted per Plantat TWo-DayIntervals over l0 Daysin Six Different H. aryridk : M. persicaeTreatrrents on CiroenhouseGrown Chile Plants,N.M.S.U,, Ias Cnrces,New Mexico,l997.

Variable Conbol Treatnent Mean Perceirtage (Dav) Meant SE Meant SE M. persicaeRdlctton TreatuentI (1:20) 2 47.88t 2.29 10.30i 2.76b 78.43 4 85.25t 2.12 12.92+.4.05b 84.85 6 118.75r 13.96 8.83r 4.63b 92.57 8 158.75r 11.66 4.92t t.46b 96.91 l0 204.75t 14.85 1.67to.72b 99.19 Treatuent2 (l:40) 2 97.t3t 3.36 17..42*2.s2b 82.06 4 141.88* 13.61 n.25 1 1.89b 91.89 6 199.88f 20.33 6.33I 1.66b 99.96 8 2s1.25tr9.09 4.33t 0.63b 98.31 l0 3lt.r3 t 2.65 1.75t 0.25b 99.43 Treatnent3 (1:80) 2 141.75X t.74 23.25tO.gob 83.59 4 200.38Xt.24 12.67t43ob 93.67 6 280.00*t2.73 7.75 t 0.66b 97.23 8 347,88rtt.49 5.08r 0.72b 98.53 l0 4n.00114.85 2.OOr 0,90b 99.52 Treatuent4 (1:160) 2 268.88r15.38 53,08r 4.07b 80.25 4 336.38r 5.48 31.67r 1.38b 90.58 6 399.88r 5.48 22.17t1.52b 94.45 8 485.8818.66 t6.25t2.25b 96.65 l0 574.s0t29.34 6.42t 1,84b 98.88 Treatue,nt5 (1:320) 2 486.38r 15.03 t66.92l,12,49b 65,67 4 590.75X27.58 94.75l-L6.O2b 83.96 6 667.88i s8.51 64.33LtO.62b 90.36 8 759.88r 50.73 50.17r 4.80b 93.39 l0 829.38r 38.01 38,50r 7.19b 95.35 Treatuent6 (1:640) 2 816.88r 25.28 421,83i.64.56b 48.36 4 891.63i 0.53 372.00t54J9b 58.27 6 t022.88116,79 291.83r 58.32b 71,46 E 1108.63i 3.71 228.67t36.29b 79.37 l0 1197.63r 8.31 t71.25t 26.96b 85.69 "StandardError bdenotes significancebetween treatment and control at p 3 0.05 usinga one-wayanalysis of variance.

251 Fig. 1 summarizesthe estimateddaily meanM. persicae densitychanges in thetreafrnent and controlcages of all six treatments.Declines in aphiddensity ztre a mensureof mortality, presumablydue to predation,while increasesare a measureof aphidnatality. In Treatnentsl, 2, 3,4,5,6, theestimated consumption rate by oneI/. axyridiswas 2.5, 3.8, 2.5, 10.9, 30,1, and 64.4pa day,respectively,while the averagenumber of M. persicaeincreased at arateof 38.7, 54.3,71.0, 75.9,85.5 and 97.8 per day, respectively, in thecontrol cages.

300 E o o E c 250 E E 800 o p 2oo € 3 €. ooo CL 1501 d o 4oo z Z(t 1oo o E = - 200 50

1:40 1:80 1:t60 1:320 1:640 Treatments

FIG. 1. Mean densityof M persicaeon chile plantsafter 10 daysexposure to H. aryridis at different densities on GreenhouseGrown Chile Plants, N.M.S.U., Las Cruces, New Mexico,l997.

The greatestM. persicae conntrrption ratewas observed during the first two daysof each treatment. This may be due to the 24-hourfast imposedon H. aryridis before eachtreatnent began.Aphid reductionwas greater than 95,00% for the 1:20,l:40, l:80, 1:160,and l:320 ratiosbut fell to 85.69%at l:640 ratios,respectively, perhaps due to theovenvhelming number of M. persicaeperH. aryridis.

ACKNOWLEDGMENT

The authorswould like to thank ShaunMeeks and JoeLaRock for their technical assistance.

LITERATT]RECITED

Bosland,P. W., A. L. Bailey, and D. J. Cotter. 1997. Growing Chiles in New Mexico. N.M.S.U.Coop. Ext. Ser.Guide H-230 pp.l-5. Canard,M., Y. Seeria,and T.R. New. 1984,p.294. In W. Junk[ed]. Biologyof Chrysopidae. Kluwer Pub.Co. Boston.

252 Chapin,J.B., and V.A, Brov 1991.Harmoniaaryridis (Pallas), the third speciesofthe genusto be found in the U.S. (Coleoptera:Coccinellidae). Proc. Entomol. Soc. Wash. 93:630-635. Dent,D. 1993.Insect Pest Management. Cab lntemational Pub. Co. Oxon,U.K. 604pp. Draper,N.R., andH. Smith. 1966.Applied Regression Analysis. John Wiley and Sons.New Yorkpp.l34-141. Georghiou,G.P. 1986.The magritudeof theresistance problem, pp. 14-53.In EdwardGlass [ed]. PesticideResistance: Strategies and Tacticsfor Management.National Academy Press.Washington D.C. Gilbert, F. S. 1993.Hoverflies. Naturalist's Handbook 5. RichmondPub. Co. Slough,U.K. pp.67. Hagan,K.S., S. Bombosch, and J.A. McMurtry, 1976.The biology and impact of predators,pp. 93-142.In C.B. Huffakerand P.S. Messenger [eds.]. Theory and Practice of Biological Control. AcademicPress. New York. Hagan,K.S., and R, van den Bosch.1968, Impact of pathogens,parasites, and predators on aphids.Arur. Rev. Entomol. 13:325-383. Harris,K., andK. Maramorosch.1977. Aphids asVirus Vectors.Academic Press. New York pp.559. Hodek, l. 1973.Biolory of Coccinellidae.Czechoslovakian Academy of Science.Prague. 6opp. Huffaker,C.B., and P.S.Messenger [eds.]. 1976. Theory and Practiceof BiologicalControl. AcademicPress. New York. p.4. Kennedy,J.S., M.F. Day, andV.F. Eastop.1962. A Conspectusof Aphidsas Vectors of Plant Viruses.Commonwealth Agricultural Bureau.lpndon. 144 pp. Knodel,J.J., and E.R. Hoebeke.1996. IPM FactSheet 101.00. Cornell Coop. Ext. Ser. pp.r-7. Mackauer,M., andM.J. Way. 1976.Myzus persicae Sulz., an aphidof world importance,pp. 51-120. InY.L. Delucchi [ed.]. Studiesin Biological Conhol. CanrbridgeUnivenity Press.Cambridge. Matthews,R.E.F. 1991. Plant Virology. 3rd Ed. AcademicPress. New York. 835pp. Quaclia,F., E. Rossi,.R. Petacchi,and C. Taylor. 1993.Observations on an infestationby greenpeach aphds (Homoptera: Aphididae) on greenhousetomatoes in ltaly. J. Econ. Entomol.86:1019-1025. Timberlake,P.H. 1943.The Coccinellidaeor lady beetlesof the Koebelecollection- Part 1. HawaiianPlanter's Record. 47 :l-67.

253 DEC.2003 vol.28 NO.4 SOUTHWESTERNENTOMOLOGIST

COMPARATryE SUSCEPTIBILITYOF EUROPEANAND AFRICANIZED HONEY BEE ' ECOTYPESTO SEVERALINSECTICIDE CLASSES

P. J. Elzen,G. W. Elzen,W. Rubink

USDA-ARS Kika de la3arzasubtropical Agricultural ResearchCenter 2413F.HwY.83 - Weslaco,TX 78596'

ABSTRACT

Three insecticideclasses (pyrethroid, carbamate, and organophosphate)were and assayedfor comparative toxicity io European honey bees, Apis mellifera L. was Africanizedhoney bees, Apis meiliferascutellata Lepeletier. Ecotypeverification candidate achievedthrough utilization of eiotype-specificisoenzyme analyses. The both insecticidestested are commonly uria in crops in the geographicalranges of wa3 e*op.- and Africanized honey bees. The European honey bee ecotype beri ,ignih"*tfy more tolerant of cyiuthrin (Pfretfuoid) than the Africanized h9ne1 to ;tp., liiely due to the preceedingdecade of pyrethroiduse in managedcolonies of .ontrof in. parasitic mite, Varroa ditructor Andersonand Trueman. Both ecotypes bees were equally susceptible to malathion (orgahophophatg)ang carbofuran (carbamate)insecticides. D-iscussion is given as to developmentof relativeinsecticide tolerancein EuropeanhoneY bees.

INTRODUCTION

The Europeanhoney bee, Apis mellifera L., is the dominant honey bee ecotype occuningthe U.S.,with distributionthroughout the entire50 states.With its discovery,in Texasin-tgg0 (Sugdenand williams 1990),the Africanizedhoney bee, Apis mellifera scutellataLepelitiei, hasnow spreadthroughout the entiresouthwestem U'S' (Sheppard andSmith 20b0). Both ecotype;in theregion now foragein concert,pollinating not only agriculturalcrops in the process,but countlesswild plant speciesas well., Both ecotypes ui. no* consideredsympatric throughout the southwestemU.S., with no efforts to eradicatethe Africanized honeybee deemedfeasible. beesand In field cropping-bles-aresituafions in the southwesternU.S., Europeanhoney Africanized honey exposedto many agricultwal chemicalsduring field foraging andpollination aitivities. The exposureof honeybees to insecticidesused in thesecrops

'Hymenoptera: Apidae 2 Disclaimer: Mention of trade namesor commercialproducts in this article is solely for the purposeof providing specific information and doesnot imply recommendationor endorsementby the u.s, Deparunent of Agricultute. 255 is a major problemimplicated in colony lossesaround treated fields. Recommendations have beenmade to lessenexposure of honey beesto toxic insecticides(Atkins et al. 1981),but actualapplication practices, such a timing of sprays,are not alwaysfeasible. Many studieshave been conductedon the toxicity of insecticideson Euroiean honey bees(Johansen et al. 1957,Anderson and Atkins 1968,Stoner et al. 19g5.cox and wilson 1987,wilson et al. 1988),but nonehave addressed the comparativetoxicity of insecticidesto Eriropeanversus Africanized honey bees in the U.S. Dankaet al. (lig6) comparedtoxicity of severalinsecticides on Europeanversus Africanized honey bees in southAmerica in the mid 1980's,characterizing honey bee ecotypes based on localityof colonies. Thesetests were conductedprior to the widespreadand intensiveuse of the syntheticpyrethroid insecticide class in Europeanhoney bee coloniesfor controlof the parasitic mite varroa destructorAnderson and Truemanand prior to movementof Africanizedhoney bees into the U.s., whereits exposureto in-hivepesticide use has been minimal. Giventhese considerations ofthe widespreadpresence ofAfricanized honeybees in agriculturalareas of the southwesternU.S. and the decadeJonguse of synthetic pyrethroidinsecticide within Europeanhoney bee colonies for V. destructorcontrol, we undertookthe presentstudy to investigatecomparative toxicity of three classesof commonlyused agriculturalinsecticides on Europeanversus Africanized honey bees. Suchtoxicity data may provide indicationsof potentialpollination disruption by either €cotypein the southwestemU.S., with implicationson appropriateinsecticide selection by farmersfor usein cropswhile honeybees axe present in fields.

MATERIALS AND METHODS

Coloniesof Europeanand Africanizedhoney bees were maintainedin separate apiaries,ca. 20 miles apart,near Weslaco,Texas. Europeanhoney bee colonieswere obtainedas commercial packages produced in centralTexas and subsequently established at Weslacoon previouslyunused wax foundationand hive equipment. Europeanhoney bees from colonies containing marked and clipped queens were verified as non- Africanized, as describedbelow and used for tests,with no acaricidetreatrnent in hives prior to test initiation. SuspectedAfricanized honey bees,originally collectedfrom swarmtraps along the Rio GrandeRiver nearHidalgo, Texas, were from coloniesthat had receivedno pyrethroidacaricide heatrnent in coloniesduring the preceedingfour years. Laboratoryverification of individual colony ecotypeand laboratorytoxicity assayswere conducted in late2000- early 2001. Isozymebioassays were used to verifr the ecotypeidentification of coloniesused in tests(Nunnamker and Wilson 1981,Spivak et al. 1988). For eachcolony, 22 adult workerfemales were collectedand frozenat -70oC until processingthe following week. Sampleswere removedfrom the freezerand kept chilled as thoraceswere removed individuallyfrom eachbee. Thesewere singly homogenized in 1.5 pl eppendorftubesin distilled water and centrifugedfor 10 min at 12,000 rpm in a refrigeratedmicro centrifuge.Pre-packaged mini PHAST gelswere then loadedwith 0.4 pl of supematant per lane and separatedusing isoelectricfocusing techniques. The co-occurrenceof malate dehydrogenaseand hexokinase bands were considered verification of Africanization for that colony; simultaneousabsence of both markers, along with identification of an original marked queen at hive inspection, were considered confirmationofa Europeancolony. For each ecotype,brood was collectedfrom three to four selectedcolonies confirmed as either Europeanor Africanized, broughtto the laboratory,and placedin an

256 incubatorovernight in order to collect newly-emergedadult worker females. Emerged beesofEuropeaior Africanizedverification were collectedthe next morning,pooled as to ecotype,and then held in cagesfor 24 hours at 28oCand 50% RH before test initiation, dwing which they were freely fed 50% sugarsyrup and distilled water- Unaesthestisedbees were then individually treatedtopically on the dorsal thorax with technical grade insecticidesdissolved in acetone. Chosen insecticideswere cyfluthrin (pyretlroid), malathion (organophosphate),and carbofuran(carbamate). All technicalgiade materialswere obtainedfrom Chem ServicesInc. (West Chester,PA 19381)and were of >95% purity. Stock solutionsof serial dilutionswere preparedby dissolvingeach insecticide in HPLC-gradeacetone; both solutionsand technicalgrade materials were stored at -20oC. Individual bees were treated with 1 pl of known concentrationor with I pl of acetoneas a controlusing a pre-calibratedmicroapplicator (Burkard, Rickmansworth,UK). Ten beeseach were treated with eachconcentration or acetoneand placed as groupsin smallscreened cages. Five concentrationsplus a control treatmentwire testedfor eachinsecticide. Treated bees were held for 24 h at 28"C in darknesswith synrp and water and then assessedfor mortality. The criterion for mortalitywas inalitity to movewhen gently prodded. Modeling trials havedemonstrated that 120 beestotal, testedfor a chemicalfor dose-mortalityestimations, are sufficient to obtain accuratedose-mortality estimates (Robertson et al. 1984). In our tests, we replicatedeach insecticide trial per ecotypethree separate times, fol a total.of 180bees pooledfrom tlree-four colonies,verified as to ecotype,tested for eachinsecticide' Dose- mortality lines for each insecticidewere analyzedby probit analysisPOLO (LeOra Softwarl; in order to estimatelethal dosevalues, confidence intervals, and slopesand standardelror of m€ans of slopes. Confidenceintervals were comparedbetween ecotypesat the LDso levels; non-overlapof confidenceintervals was considered demonstrationof significant differencebetween ecotypes in responseto a particular insecticide.For eachchemical, equivalency ofslopes was determinedby non-overlapof two timesthe standarderror of the mean.

RESULTSAND DISCUSSION

Resultsof isozymeanalyses of samplebees from coloniesused in testsconfirmed the identificationofecotypes as either European or Africanizedcolonies. Thosecolonies establishedfrom swarmi collected in South Texas were clearly Africanized as demonstratedby high frequenciesof both isozymemarkers, seen upon assaycompletion' Colonies estabilstrJdfrom packageswith marked and clipped queensrevealed the absenceof the two characteristic marker isozymes, confirming their non-Africanized status. Selectedcolonies were used,as describedabove, for all subsequentinsecticide assays,ensuring the continuedverification of non-AfricanizedEuropean honey bees by thepresence of markedqueens. Both malathion and carbofiran were equally toxic to Europeanhoney beesand Africanizedhoney bees, as evidenced by overlapof 95o/oconfidence intervals at the LD56 levels (Table l). Responseto cyfluthrin was significantlydifferent betweenthe two ecotypes,with Europeanbees exhibiting a greatertolerance to this pyrethroidcompound thanAfricanizedbees (Table l). This greatertolerance by Europeanhoney bees may be dueto the preceedingdecade ofpyrethroid usein managedEuropean colonies to control the honeybee parasitic mite V. destructor.Because V. destructorhas been considered to have severely reduced feral populationsof the Europeanhonby bee in the U.S. (Hoopingameret al. 1992), selection pressure for pyrethroidtolerance in Eruopeanhoney beesiikely occurred with little infusion of unexposedsusceptible phenotypes from the wild.

257 TABLE 1. Responseof Europeanand AfricanizedHoney Beesto TopicalryApplied Insecticides;Values Given areExpressed as pg Insecticideper Bee.

Treatment Ecotype n LDso(95% CD" LDeo(95% CD slope(+SEM)

Malathion AHB 180 0.123(0.093-0,190) 0.279(0.183-0.706) 3.602 (0.6s2) EHB 180 0.193(0.167-0.218) 0.292(0.2s4-0.37r) 7.1 14 (1.184)

Cyfluthrin AHB 180 0,01I (0.008-0.018)*0.030(0.024-0.043) * 3.109 (0.555) EHB 180 0.036(0.020-0.051) 0.070(0.0s0-0.202) 4.s2r(0.669)

CarboturanAHB 180 0.033(0.027-0.038) 0.057(0.048-0.076) 5.400(0.974) EHB 180 0.039(0.034-0.045) 0.062 (0.053-0.082) 6.388(1.132) " Values followed by an asteriskare significantlydifferent for that insecticide betweenecotypes.

In conhast,Africanized honey bees are rarely maintained in managedcolonies in the southwesternU.S., with little intenseselection pressure by pyrethroidpesticides to control z destructoi,opposite to what is experiencedby Europeanhoney bee colonies, Slope values for each insecticidetested, comparing ecotypes, were rot significantly different, indicating similar homogeneityof responseby both ecotypes to eaeh insecticide. In testing carbamateand organophosphatesusceptibility after eight yearsof observations,Anderson and Atkins (1968) found no changein toieranceby european honeybees to theseinsecticide classes. Tucker ( I 980),however, was able to demonstrate increasedtolerance ofworker Europeanhoney bees after selectionfor nine generations _ with carbaryl. Dankaet al. (1986)found in their studiesofAfricanized and European honey bees in Venezuela,with locationof swarmsused as designationof Africanizedcolonies, that Africanizedhoney bees were generallymore tolerant than Europeanhoney bees toward th'reeof the four classestested, results they found surprisinggiven the smallersize of Africanizedhoney bees. They discussthe role of degreeof cuticularsclerotization as a possibleexplanation for susceptibilitydifferences; in their study, they used newly emergedbees that wereless than 24-h old. In our toxicity trials,however, we usedadults that were at least 24-h old after emergenceand so perhapswere more sclerotizedthan adultsused by Danka et al (1986). Also, in our studies,we used populationsof Europeanand Africanized honey bees that likely receivedsimilar exposureto agricultural chemicalsduring the preceedinggenerational years, other than pyrethroidsused in managedEuropean colonies, and so may explain the greatersimilarity of responseto carbamateand organophosphateinsecticides tested in our study. The additionalstep of definitively confirming colonies as Africanized or European,as we did in our study,may alsoexplain the differencesin responseswe saw comparedto thoseseen by Dankaet al (1e86) In the southwestemU.S., the primary ecotypefound existingas feral coloniesis the Africanizedhoney bee, with Europeanhoney bees virtually absentfrom the wild. This appearsto be due, however,to differentialincreased ability of Africanizedhoney beesto withstandinfestations of V. destructor@ubink et ^l 1995),while Europeanhoney bees are more severely affected and colonies are rapidly killed by v. destructor

258 infestations. There are no field data to indicate that the greater presenceof feral Africanized honey beesin the southwesternU.S., as well as in much of Latin America, comparedto Europeanhoney bees,is due to any greatertolerance of the Africanized ecotypeto insecticides. It is also important to state that, while statistically significant differenceswere seenbetween ecotypes in our study,as well as in Dankaet al (1986), such differencesin dose-mortalityvalues would be minor comparedto the much higher ratesof insecticideuse in field pest control settings. Insecticidesare usedin kilogram quantitiesper hectare,rates most likely much higher than what we used to determine dose-mortalityestimates in our study. In order for selection for ecotype resistanceto occurin a field setting,there must obviouslybe mortalityof susceptibleindividuals, but also a concomitant survival of tolerant individuals. The insecticide rates used in freld crop settingsto which eitherecotype would be exposedwould seemnot to allow survival of slightly tolerant individuals due to the excessiverates bees would experience.It is also important to note that differentials in Europeanhoney bee reproductivefitness has been demonstratedfor Europeanhoney beesexposed to fluvalinate (pyrethroid) pesticideused in managedcolonies to control V. destructor'.Rinderer et al (1999) demonstratedthat drones exposed to fluvalinate in managedcolonies at normal usage rates lived significantly shorter than non-exposed drones and had minor reproductive fitness disadvantages(i.e., clear selection pressurefor tolerance) and thus allowed for "selection" by queansof potentially more tolerant dronesfor productionof offspring with resultant greater tolerance to pyrethroid pesticides. Such selective advantageby potentially tolerant dronesmay explain the greatertolerance of Europeanhoney beesto the pyrethroidinsecticide tested in our studycompared to Africanizedhoney bees in our study that did not receive such a level ofselection within colonies. Additionally, only non-reproductiveworker beesare typically exposedto high dosesof field agricultural insecticides, possibly explaining further why resistance to organophosphateand carbamateinsecticides has not arisenin either ecotype. The reasonthat Europeanhoney beeshave not developedresistance to organophosphatesin particular, given the usein the past 5 yearsof an acaricidein this classin managedhives, is unknown. As usage of this class,and potentially of other classesof pesticideswithin hives continues,managed Europeanhoney bees may exhibit increasedtolerance to these compounds. Because developmentof pesticideresistance in organismsis difficult to predict timewise,only continuedtesting will reveal if suchresistance will actually occur'

ACKNOWLEDGEMENT

We wish to thankNoe Buenrostro,Jesus Maldonado, and Roy Medranofor technicalassistance.

LITERATURE CITED

Anderson,L. D., and E. L. Atkins, Jr. 1968. Pesticideusage in relationto beekeeping. Ann, Rev.Entomol, 13l.213-238. Atkins, E. L., D. Kellum, and w. Atkins. 1981. Reducingpesticide hazards to honey bees: mortality prediction techniques and integrated managementstrategies' Univ.Calif. Div. Agric.Sci. Leafl.2883. cox. R. L. andw. T. wilson. 1987. Thebehavior of insecticide-exposedhoney bees. Amer.Bee. J. 127:ll8-119.

259 Danka,R., T.E. Rinderer, R. L. Hellmich II, and A. M. Collins. 1986.Comparative toxicitiesoffour topicallyapplied insecticides to Africanizedand European honey bees(Hymenoptera: Apidae). J. Econ.Entomol. 79: 18-21. Hoopingamer,R. A., andG. D. Waller. 1992. CropPollination, pp. 1043-1082. In Joe Graham[ed.] The Hive andthe HoneyBee. Dadant Press, Hamilton, IL. Johansen,c. A., M. D. coffey, andJ. A. Quist. 1957. Effect of insecticidetreatments to alfalfa on honeybees, including insecticidal residues and honeyflavor analyses. J. Econ.Entomol. 50:721-723. LeOra Software. 1987. POLO-PC:a user's guide to probit or logit analysis. LeOra Software,Berkeley, CA. Nunamaker,R. A., and W. T. Wilson. 1981. Comparisonof malic dehydrogenase allozymepatt€rns in the Africanizedhoney bee (Apis melliferaadamsonii L.) and the Africanizedpopulation of Brazil. J. Kan.Entomol. Soc. 54:704-710. Rinderer,T. E., L. I. de Guzman,V. A. Lancaster,G. T. Delatte,and J. A. Selzer. 1999. Varroa in the mating yard: I. the effects of Varroa jacobsoni and Apistan on dronehoney bees. Amer. Bee J. 139:134- 139. Robertson,J. L., K. C. Smith,N. E. Savin,and R. L. Lavigne. 1984. Effectsof dose selectionand sample size on the precisionof lethal dose estimatesin dose- mortalityregression. J. Econ.Entomol. 77:833-837. Rubink, W. L., W. T. Wilson, A. M. Collins, and P. Levano-Martinez. 1995. ComparativeAfricanization rates offeral honeybee populations at threelatitudes in northeastemMexico andsouthem Texas. Amer. Bee J. 135:829. Sheppard,W. S., and D. H. Smith. 2000. Identificationof African-derivedbees in the Americas:A surveyof methods.Ann. Entomol.Soc. Amer. 93: 159-176. Spivak,M., T. Ranker,O. Taylor, W. Taylor,and L. Davis. 1988. Discriminationof Africanized honey bees using behavior, cell size, morphometrics,and a newly discoveredisozyme polymorphism, pp.313-324. 1n G.R.Needham, R.E. Page, "African" M. Delfino-Baker, and C. E. Bowman [eds.], The honey bee. Westview,Boulder, CO. Stoner,A., W. T. Wilson,and J. Harvey. 1985. Acephate(Orthene@): effects on honey beequeen, brood and worker survival. Amer.Bee J.125:448450. Sugden,E. A., andK. R. Williams. 1990. October15, the day the beearrived. Glean. BeeCult. I 19:l8-21 . Tucker,K. W. 1980. Toleranceto carbarylin honeybees increased by selection.Amer. BeeJ. 120:36-46.

260 DEC.2003 vol".28 NO.4 SOUTHWESTERNENTOMOLOGIST

SUTTABTLITvOF FORMIC ACID TO CONTROLVARROADES?RUCTORi AND SAFETYTO APIS MELLIFEM2IN THE SOUTHWESTERNU.S.3

P. J.Elzen

USDA-ARS Kika de laGarzasubtropical Agricultural ResearchCenter 2413F,.Hwy.83 Weslaco,TX 78596

ABSTRACT

Formic acid in a gel form was testedin south Texasfor its efficacy in controlling the parasitichoney bee rnite, Varroa destructorAnderson and Trueman,and its safetyfor use in southemconditions. Formic acid gel gave significantly better control than untreatedcolonies, but such control was significantly less than standardtreatments of fluvalinateand coumaphos. The formic acid gel treatmentalso resulted in significant,but not excessive,mortality of adult honey beescompared to untreatedcontrols' This was particularlyseen €rs temperatures consistently rose above ca.24"C. Queenmortality was also observedin coloniis treatedwith formic acid. Discussionis given on integrationof the formic acid gel pack into a V. destructormanagement plan for southwestemU.S. beekeepers.

INTRODUCTION

The varroa mite. Varroa destructorAndersen and Trueman,is one of the most seriousparasitic pests attackinghoney bees in the U.S, and on a worldwide basis (shiman-ukiet al.1gg2). This mite damagesimmature and adult honeybees by feeding on bee hemolymph and by transmittingviral pathogens. Left unchecked,a heavy infestationof V. destluctor can weakenand kill an entire colony within ca' two yearsand causeeventual loss of mostcolonies within an apiary. The primary meansof controllingZ destructorsince its discoveryin the U.S. in 1987(Anonymous 1987) has been through synthetic chemical acaricides. The pyrethroid fluvalinate (Apistan@)and the organophosphatecoumaphos (CheckMite+@) have been the mostcommonly used EPA-approved compounds for Z destructorcontrol in the U.S. Beginningin 1998,however, U.S. beekeepersin widely diverseregions began noticing decreasedefficacy of fluvalinate in their operations,as evidencedby unexpectedlyhigh mite numbersin treatedhives. Suchdecreased efficacy was confirmedas resistanceby Y. destructorto this compoundby Elzen et al. (1999). Beekeepersthen tumed to coumaphos,which providedexcellent control of fluvalinate-resistantmites (Elzenet al.

' Mesostigmata:Vanoidae 'Hymenoptera: Apidae 3 Mention of a trade name does not constitute endorsementby the USDA 261 2000). Elzen and westervelt (2002), however,confirmed that v. destrucforhas now becomeresistant to coumaphosin Florida,with likely spreadto otherregions of the U.s. asbee colonies are moved as part of normalmigratory activities of beekeeping. Beekeepershave now increasingly called for the availability of altemative compoundsfor the control of v. destructor. one such compoundis formic acid, which hasreceived approval for useas a gel formulationby the U.S. EnvironmentalProtection Agency. Researchon the suitability of formic acid liquid and gel formulationshas been conductedin the U.S., but almost exclusivelyin the northern regions of the U.S. (Feldlauferet al. 1997,Calderone and Nasr 1999,Calderone 2000). Beekeepersin the southernU.S. have called for information on the suitability of formic acid for Z destructorcontrol in the relativelywanner conditions ofthe southand associatedbffects on honey bees in treated colonies. We thus initiated the present study to provide beekeepersin the southwestemU.S. with datathey would find necessaryto considerthe useof formic acid in their operations.Because the only EPA-approvedusage of formic acid registeredcurrently in the U.S. is the gel formulation,we choseto evaluatethis formulation in conditions of the southwesternenvironment.

MATERIALS AND METHODS

In order to ensure the highest contact of gel pack vapors with mites, treattnent was initiated when the majority of V, destructor were on adult bees, where formic acid vapors would most readily reach phoretic adults. For this reason, the trial was initiated on 3 January 2002, a general date during which colonies in south Texas are most broodless (<5 frames of brood per colony). Colonies were located at Weslaco, Texas, at a site at least 5 miles from other apiaries. Colonies were established in the spring of 2001 from commercially-obtained packages, installed in new equipmenVfoundation, and remained untreated for V. destructor until initiation of the present study. All colonies were queenright and contained ample stores of honey and consisted of a deep bottom brood chamber and one deep honey super. Standard compounds consisted of l0% fluvalinate strips (Apistan@, obtained from Dadant; Paris, TX) and l0% coumaphos strips (CheckMite+@, obtained from Dadant; Paris, TX). In both cases,one strip of either treatment was inserted in the bottom brood chamber, as recommended by manufacturers (one strip per five frames of brood). The formic acid treatment consisted of 200 g of formic acid gel (obtained from Befter Bee; Greenwich, NY) in vegetable baggies, each of which was cut open in a X configuration immediately before application. One pack was placed over the end of the top bars of the brood chamber, but not over the brood area. One-inch spacers were inserted between bottom and top chambers,to allow for air flow. Spacers were inserted in all other treatment and control colonies as well. The control treatrnent consisted of colonies receiving no strips or gel packs. Treatments were randomly assigned to colonies, with five to six colonies per treatment. Effects of treatments on V. destruclor control were assessed according to the method of Feldlaufer et al. (1997). Sticky boards, under screens, were inserted on the bottom board of each colony at the start of the trial. Treatments were then immediately applied. Boards were removed weekly and numbers of fallen mites were counted for that week. This was repeated for a total of three weeks and was considered the treatment period. A following 3-week evaluation was then begun in all colonies by removing previous treatments and inserting one coumaphos strip in the brood chamber as a final clean-up treatrnent. Weekly mite drop was assessedas described for the treatment phase. Final efficacy of treatments was calculated by summing the number of mites dropped

262 nynler dropped during the heatment and evaluation phases(total drop), dividing the 100. These J*in! trr. treatmentphase by the total drop, and multiplyingthis value by and p"r""i"g" efficacy nulo", *"r" arc-sine tansformed and analyzed by ANOVA meansseparation tests (LSD, P:0.05). positioning Effect of treatmentson adult honey bee mortality was assessedby phase and standard dead bee traps at colony entrancesduring both the treatment and were .uotu"tion prr^". Numbers of dead bees were counted at 3-day intervals treatmentwere i.rou.a from traps. For eachdate, mean numbers of deadbees for each compared-- by ANOVA andmeans separation tests (LSD, P=0'05)' ' website T.-pout*es during the trii were obtainedfrom a local weatherstation in deadbee traps service. Final queenstatui was assessedby noting any deadqueens during the treatrnentand evaluationphases.

RESULTSAND DISCUSSION

comparedto All treatmentsapplied for V. destructorprovided significant control gel pack,however, untreatedcolony resultilf*t. t;. The effrcacyof the formic acid There *^ .igrin.*tiy less than the standardtreafinents of fluvalinate or coumaphos' and as wasalJo wide variationin controlwith formic acid,with as low as 69'8%control affordedby higft ^ S:.gX control for individualcolonies. In contrast,ranges ofcontrol providing consistent fluvalinateand for coumaphoswere much nEtrrowef,with both regions of the "ont of g."ur"r than ca. 90t%. However, such consistentcontrol in other of resistance U.i. Uyifuvatinateand now coumaphosmay be.questionable,due to reports 2002)' Our iy V. irrtru"tor to rhesecompounds (Elzen et al. 1999,Elzen and Westervelt than the 70'3% resultsof overall 84.5%"onrot with the formic acid gel pack is higher in the spring .ont of t"pota"d by Fetdlauferet al. (1997) for tests run in Maryland individual months. They alsoreported relatively high variability in mite control between colonies,as we observedin oqr study.

Fluvalinate 95.0(1.1) a 90.4-97.7 Coumaphos 95.8(1.4) a 90.8-98.8 FormicAcid Gel 84.s(4.7) b 69.8-93.8 Control 33.8(7.4) c I I .7-54.1 oValueswithin a column followed by different letters are significantly (ANOVA,LSD,P<0.05).

Statistically significant mortality of adult bees in the formic acid treated colonies was observed, compared to untreated control results (Table 2), starting at the sixth day of the treatment perioi. While significant, mortality was not excessive. Prior to this time, the temperature maximum averaged 15.5"C. Temperaturesafter day 6 averaged 24.loc. No significant bee mortality was seen for fluvalinate or coumaphos treated colonies, compared to unteated colony mortality, after the sixth day of treatment and warmer

263 I RI 5;ssl 6t geesdl ssFsldl 9 \o ool I !l (g cd ^ (gl ^^N ^l 6 ?9r; ql \O € - 6l (-- vvvvl KssKlc'].! q nl

€..o.lYl3 QcQcll H .;

a=Aal€ I vl/vvl=c.l -j o\l o $eQnl€ s Fx; NIE te lc

O €E!l'!

o F EEEsIs 264 with 20% temperatures.Queen mortality was also seen,inthe formic acid replicates, phaseof the trial' queen' mortality observedduring the treatment Because beekeepersare facing loss of both fluvalinate and coumaphos as The efficacious acaricides,there is naturalfu great interest in alternative compounds" in iormi" u"ia gel pack would be a logical choice,but our resultsshow that variability control fronihive to hive in southwestemU.S. conditions would have to be tolerated. Suchvariability would limit the attractivenessof the gel packin standardU.S. bee stocks, Untlnuy be significantly less important in newly developedbee stocks_thatexhibit toleranceto V. destructorinfestations (Rinderer et aI.2001, Harbo and Harris 2001' spi"ar< uno Reuter 2001). These stocks are capableof withstandinghigher mite The issueof w91ke1and queen 'tpopulations, with only occasionaltreatnent necessary. iti woUa be of considerationfor southernbeekeepers, where unpredictablefall and wintertemperatures may increaseand so havepotentially negative effects in formic acid gel packtreated coloniei. An alternativesolution to this possibilig at leastfor migratory 6."i.""p"rr, would be to apply treatnents while colonies may be in cooler northem locations,before migratingio southern(and warmer)locations. Integrationof tolerant beestocks, colony loiation, andthe formic acidgel packmay thusbe a workablesolution in dealing with Z destrnctor that are rapidly developing resistanceto all standard fteatrnentsfor its control.

ACKNOWLEDGMENT

I wish to thank Noe Buenrostro and Jesus Maldonado for their technical assistance.

LITERATURE CITED

Anonymous.1987. Vanoa mitesfound in the united States.Amer. Bee J. 127:745- 746. Calderone,N. W. 2000. Effectivefall treatmentof Varroajacobsoni (Acui: Varroidae) with a new formulation of formic acid in colonies of Apis mellifera (Hymenoptera:Apidae) in the northeasternUnited states. J. Econ.Entomol. 93: 1065-1075. Calderone,N. W., andM. E. Nasr. 1999. Evaluationof a formic acidformulation for the fall control of Varroajacobsoni (Acari: Vanoidae)in coloniesof the honeybee Apis mellifera(Hymenoptera: Apidae) in a temperateclimate. J. Econ.Entomol. 92:526-533. Elzen. P. J., F. A. Eischen,J. R. Baxter, G. W. Elzen, and W. T' Wilson' 1999' Detectionof resistancein U.S. Varroaiacobsoni (Mesostigmata:Vanoidae) to theacaricide fluvalinate. Apidologie 30: 13-17. Elzen,P. J., J. Baxter,M. Spivak,and w. T. wilson. 2000. control of vanoa jacobsoni Oud.resistant to fluvalinateand amitrazusing coumaphos. Apidologie 3l: 437- 44t. Elzen,P. J., and D. Westervelt. 2002. Detectionof coumaphosresistance in Varroa destructorin Florida. Amer.Bee J. 142:291-292. Feldlaufer,M. F., J. S. Pettis, J. P' Kochansky,and H. Shimanuki' 1997' A gel formulationof formic acid for the controlof parasiticmites of honeybees, Amer. BeeJ. 137:661'663.

265 Harbo,J. R., andJ. w. Hanis. 2001. Resistanceto varroa destructor(Mesostigmata: Vanoidae) when mite-resistantqueen honey bees(Hymenoptera: Apidae) were free-matedwith unselecteddrones. J, Econ.Entomol. 94: 1319-1323. Rinderer,T., L. de Guzman,G. Delatte,J. Selzer,J. Williams, L. Beaman,V. Kuznetsov, M. Bigalk, S. Bemard, and H. Tubbs. 2001. Multi-state field trials of ARS Russianhoney bees l. Responsesto Varroa destructor1999,2000. Am. Bee J. l4l:658-661. Shimanuki,H., D. A. Knox, B. Furgal4D. M. Caron,and J. L. Williams. 1992.Diseases and pests of honey bees, 1n The Hive and Honey Bee. Dadant and Sons. Hamilton,IL. Pp 1083-ll5l. Spivak,M., andG. Reuter. 2001. Varroadestructor infestation in untreatedhoney bee (Hymenoptera:Apidae) colonies selectedfor hygienic behavior. J. Econ. Entomol.94:326-331.

266 vol.28 NO.4 SOUTHWESTERNENTOMOLOGIST DEC.2003

IDENTIFICATION OF THREEFORENSICALLY IMPORTANT BLOW FLY (DIPTERA:CALLIPHONDAE)SPECIES IN CENTRAL TEXAS USING MITOCHONDRIAL DNA

FelixM. Tenoria",Jimmy K. Oslonb,and Craig J' Coatsc TexasAUM University,Department of Entomology CollegeStation, Texas 77843

ABSTRACT

Themorphologicalidentificationofthreeblowflyspeciescommon-tocentral Texas,cynomyopsis iadaveriza (Townsend),calliphora livida (11all), and calliphora of PCR ii"in; was'confirmedusing restriction enzyme analysis lioAineau-Desvoidy), species amplifiedmtDNA ftugm#r. An amplifiedregion of the COI genefrom each pattems was subjectedto digestionby four restrictionenzymes. Restriction fragment each distinguishedthe thrie speciei.The DNA sequenceof the COI genefragment from rp".i.". was determined,confirming the ristriction enzyme analysis'.subsequently, to .tlNe fragmentsfrom field collicted sampleswere amplifiea q9 subjected of the restriction Jnryrrr" analysis in order to test the validity and reliability identificationmethod. For each speciestested, it was demonstratedthat restriction different enzymedigestion of amplified .toNe fragmentscould be usedto identify stages' blow fly dJvelopmentalslages to species,previously a problematictask for early whereonly limited morphologicalcharacters are available'

INTRODUCTION

Eachblowflyspecies(Diptera:Calliphoridae)hasaparticulardevelopmental time. This makes them usefui for forensii investigations,provided that collected samplescan be correctlyidentified to species.However' speciesidentification using morphologicalcharacteri can be problematic,especially in earlydevelopmental stages' tnis is pirticularly relevant,as in many instancesinsect evidence collected at a crime ,".n" i, predominantly eggs and larvae. The larvae of some species are indistinguis-hable,especiall/-the early instars (Debang and Greenbe^rg1989)' Furthermore,to preservespecimens at the time of collection.they are often killed' preventingthe specimensfrom being rearedto adulthoodfor identification. Eggs pose an evengreater problem as the morphologiesofmany speciesare indistinguishable and arealso Jusceptibte to desiccation.o tttutitt. embryomay die beforeeclosion, making it impossibleto determinethe species.Improper species identification could lead to an

" Correspondingauthor. Tel. (214) 215'5627; e-mail: mariana-tenorio@hornail'com' b Depart-ent oiEntomology, TexasA& M University,College Station,Texas 78743 " DepartrnentofEntomology, TexasA& M University' CollegeStation, Texas 78743

267 inconect estimate of the post mortem interval (PMI). Therefore, the ability to perform a molecular identification from any insect evidenceis critical. "fingerprinting", DNA identification, or DNA has recently been applied to the identification of insects of forensic importance (Black and Duteau 1997, Roehrdanz and Johnson 1996, Sperling et al. 1994, Vincent et al. 2000, Wells and Sperling 2001, Azeredo-Espin and Madeira 1996, wallman and Donnellan 2001)); Mitochondrial DNA (mtDNA) is relatively resistant to degradation and is particularly well suited as a template for identification purposes. There are hundreds of mtDNA copies in each cell and the mutation rate of mtDNA is high enough to provide numerous sequence differences among closely related species (Sperling et al. 1994, Lee and Gaensslen 1990). Most importantly, mtDNA can be used for the identification of all insect life stages.It is important that tests and confirmations of molecular identification methods are performed for each geographical region. This will help overcome possible complications such as local fixation of particular polymorphisms, and enable the molecular evidence to be used more effectively in a local court of law. Four blow fly species were chosen for this study, three "bluebottle" species, Calliphora vicina (Robineau-Desvoidy), Calliphora livida (Hall), and Cynomyopsis (=Cynomya) cadaverina (Townsend), and the black blow fly, Phormia regina (Meigen). The occurrence ofthese specieswas recorded during a field study conducted in College Station, TX and a rural study site near College Station, TX. The "bluebottle" species chosen for molecular identification were the only ones collected of those that have been recorded for the area. The prevalence of the bluebottle species dur:ng the winter and spring months in Brazos and Burleson Counties of central Texas and the differences in their developmental time makes these species perfect candidates for forensic applications. In addition to general difficulties with adult and pupal morphological identification keys, the identity of C. vicina md C. livida larvae cannot be resolved (Debang and Greenberg 1989). The black blow fly, Phormia regina (Meigen), was chosen for comparison becausethis specieshas been used in a previous mtDNA-based species identification study (Sperling et al. 1994). Sperling et al. (1994) PCR amplified a region of the COI mitochondrial DNA gene and digested the PCR product with four restriction erzymes to establish a method for distinguishing between three blowfly species, Phormia regina, Phaenicia sericata, ard Lucilia illwtris. Applyng the same restriction enzymes utilized in the previous study (Sperling et al. 1994), we PCR amplified and digested a 348 base pair sequence of the COI mitochondrial DNA gene and identified restriction fragments on agarose gels. The results ofthese studies are presented,as well as an analysis ofthe DNA sequenceofthis generegion from the three bluebottle species.

MATERIALS AND METHODS

Individualwild caughtfresh blow fly specimensfrom collectionsmade at two field studysites (Texas A&M MosquitoResearch Laboratory, College Station, Brazos County,Texas; 4km SSW Snook,Burleson County, Texas) were killed by snapfreezing in liquid nitrogenand DNA was extractedfollowing Black andDuteau (1997). Only the thoraxof eachadult specimenwas usedfor DNA extraction,the head,abdomen and wingswere saved as vouchers for morphologicalidentification. A regionof the mitochondrialgenome within the cytochromeoxidase subunit I gene (CO! was amplified using the PolymeraseChain Reaction(PCR). The PCR primersused in this reactionare described by Sperlinget al. (1994)and amplification of a 348bp gene fragment.PCR was performedin an MJ Research@PTC-200 Peltier ThermalCycler (MJ Research,Inc. Waltham,MA 02451).Ten microliters(pl) of DNA

268 from each specimenwas used in a 40pl total reaction volume (0.5p1 Taq DNA polymerase[Promeg4 Madison, WI 53711-5399],4.0p1 thermophilic magnesium free lOX buffer [Promega,],4.0p1 25mM MgCl2* Buffer, 4.0prl 2.5mM dNTP mix [Promega],20pmoles COI (5') primer, 20pmolesCOI (3') primer, and ddH20 to 40pl per sample).A preincubationat 94oC for 5 min. was followed by 30 cycles of denaturationat 94"C for 30 seconds,annealing at 45oCfor 30 sec.,and extensionat 72"C for 30 seconds.The procedurewas terminated with a final extensionat'12"C for 5 minutes. For restrictionfragment analysis, 10pl of the PCR productswere digestedin a 20p1total volume accordingto the directionsspecified by Promega.The resulting productswere separatedon a 2%oagarose gel and imaged using a UVP BioDoc-It system(UVP, Cambridge,UK). For DNA sequenceanalysis, the PCR productswere gel purified and cloned using the pGEM@-T Easy Vector System I (Promega) accordingto the instructionsprovided by the manufacturer.Miniprep DNA was preparedaccording to the inshuctionsprovided with the QIAprep@Spin Miniprep Kit (QIAGEN hc., Valencia,CA 91355).Following DNA purification,5pl reactions(2pl Big Dye Mix, l0pM of primer,4OOngramsof DNA template,and ddH20to 5pl) were preparedfor DNA sequencingfollowing the ABI PRISMTMBigDyerM Terminator Cycle SequencingReady Reaction Kit (PE AppliedBiosystems, Alameda, CA 94501).After a preincubationat94"C for 3 minutes,and 30 cyclesof94oC for 30 seconds,50oC for 30 seconds,and 60'C for 1 minute,each reaction was run througha Micro Bio-Spin@P-30 Tris Chromatography Column (BIO-RAD, Hercules, CA 94547) to remove unincorporatednucleotides. Samples were dried down and submittedto the Gene TechnologiesLaboratory (Texas A&M University) for processing. DNA sequence analysisand restriction enzyme mapping was performed with the VectorNTI@ Suiteof programs(Informax@, North Bethesda,MD 20852).

RESULTSAND DISCUSSION

Fig. I showsthe pattem of restrictionfragments generated by four different enzpe digestionsof the COI mtDNA gene fragment for each species.Species identificationcan be made with the restriction enzpe digestion.EcoRV does not distinguishbetween any of the species.The DdeI pattem identifies C. livida from the rest of the speciesand the DraI pattemidentifies P. regina. Last|y, theHin/I pattem can distinguishbetween C. cadaverinaand C. vicina(Table l). ClonedCOI PCR productsfrom two additionalindividuals of eachof the three bluebottle specieswere subjectedto DNA sequenceanalysis. Fig. 2 shows the alignment of one DNA sequencefrom each species. While single base pair polymorphismswere identified between mernbers of the samespecies (data not shown), thesechanges did not occur at the chosenrestriction enzyme recognition sites. ln a previousstudy by Vincent et al. (2000), amplified mtDNA from the COI gene of Phaeniciaspp. specimens was analyzedby restrictionenzyme digestion to determineif the restriction enzymesused in the study by Sperling at al. (1994) correctly identify species.In that study,samples sizes were small (n:3,4, or 5), and the resultsshowed that polym.orphismsin membersof the samegenus and speciesmade the specimens indistinguishablefrom eachother. Here, PCR amplifiedproducts from the col geneof a further 23 individuals of each specieswere subjectedto digestionby all four restriction enzymes.Identical restriction fragment pattems were observedfor all individualswithin eachspecies, for all enzymestested (data not shown).Egg andlarval

269

C, cadaveri-na c. Iiviala C. vicina

C. cadaveriDa C. livida C, vicida

C. cadaverina 121 C. Iiwida L2L C. vicina l2L

cadaveTina 181 CTGTATAATACCGAATACAGCTCCTATAC},TAGCACATAATGGAAATGGGCAACTACAT Livida 181 .. . . .A...... AT.....c.. . c vicina 181 ..,...... AT,

c cadaverina 241 c livida 24L vicila 24L

C. CAdAVETiNA 3OT TTCTACTGTAAATAAGAATACAAACCCTAAAGCICATA.AAGTAGCTCI C. Livida 301 . .C...... c....cA... C, vicina 301 ..C.

FIG. 2. Sequencedifferences within the 348bpCOI genefragment of eachbluebottle species(GenBank Accession Numbers AY036971 - C. cadaverina; AY036969- C. livida; AY036970 - C. vicina). Only sequencedifferences are shown. Restriction enzpe recognitionsites are highlighted with boxes.

In conclusion,the mtDNA identificationprocedure used in this study was reliablefor positively identiffing the four speciestested. However, the resultsof this studyshould not be usedas a definitivetool for identificationwithout first examining all availablemorphological characteristics as well as geographicand seasonaldata. without taking into accountthese other sowcesof information,it may be possibleto make a misidentification, particularly if further suweys reveal speries that share the sameor similarrestriction enzyme digestion profiles for this mtDNA region. More surveys and experimental testing must be performed to enable other closely related speciesof blow flies occurring in the central rexas area to be distinguished.The genusPhaenicia is an excellentcandidate for further validationof thesemethods. In a recentfield studyperformed in centralrexas (Tenorioet al. 2003), two speciesfrom this genuswere cataloguedin the samegeographical area. These speciesare closelyrelated and sharemany of the samemorphological characteristics. They typically occur during different seasonsof the year but could easily be misidentifiedbecause of a lack of suitable keys for identification purposes.For example, some members of the Phaenicia genus occurring in Texas are indistinguishableeven as adults. Further researchwill assist the validation of an entomologicalapproach to the estimationof postmortem intervals, geographic location ofdeath,and recent history ofthe individual.

27r ACKNOWLEDGMENT

We thankMichelle McNeil for her assistancewith fly specimencollection, Chad Smith and David Pledgerfor technicalassistance, and SpencerJohnston for critical commenB.

LITERATURE CITED

Azeredo-Espin,A.M.L., and N.G. Madeira. 1996. Primarymyiasis in dog causedby Phaeniciaeximia (Diptera:Calliphoridae)and preliminarymitochondrial DNA analysisof thisspecies in Brazil. J. Med.Entomol. 33:839-843. Black, W.C,, and N.M, Duteau. 1997. RAPD-PCRand SSCPAnalysis for Insect PopulationGenetic Studies, pp. 361-373. In MolecularBiology of Insect DiseaseVectors: A MethodsManual. Chapmanand Hall, New York, NY. Debang,L., and B. Greenberg. 1989. Immaturestages of some flies of forensic importance.Ann. Entomol.Soc. Amer. 82:80-93. Lee. H.C., and R.E. Gaensslen. 1990. DNA and Other Polymorphismsin Forensic Science.Year Book MedicalPublishers, Inc., Chicago,IL. Roehrdanz,R.L., and D.A. Johnson.1996. MitochondrialDNA restrictionsite mapof Cochliomyiamacellaria (Diptera:Calliphoridae). J. Med. Entomol.33:863-865. Sperling,F.A.H., G.S.Anderson, and D.A. Hickey. 1994. A DNA-basedapproach to the identificationof insect speciesused for postmorteninterval estimation. J. ForensicSci. 39:41 8-427. Tenorio,F.M., J.K. Olson,and C.J. Coats. 2003. Decompositionstudies, with a catalog and descriptionsof forensicallyimportant blow flies (Diptera:Calliphoridae)in centralTexas. Southwest.Entomol. 28:37-45. Vincent,S., J.M. Vian, andM.P. Carlotti. 2000. Partialsequencing of the cytochrome oxidaseb subunitgene I: A tool for the identificationof Europeanspecies of blow fliesfor postmorteminterval estimation. J' ForensicSci. 45:820-823. Wells, J.D., and F.A. Sperling. 2001. DNA-basedidentification of forensically importantChrysominae (Diptera:Calliphoridae). J. Forensic Sci. Int. 120:110- I 15. Wallman,J.F., and S.C. Donnellan. 2001. The utility of mitochondrialDNA sequences for the identificationof forensicallyimportant blow flies (Diptera:Calliphoridae) in southeasternAustralia. J. ForensicSci. Int. 120:60-67

272 vol.28 NO.4 SOUTHWESTERNENTOMOLOGIST DEC.2003

USING COLD-STOREDOR OVERWINTERINGI CERIA MALH E R BA E NUZZACT (ACARINA: ERIOPHYIIDAE),A GALL-FORMINGERIOPHYIID MITE. FoR TNFESTATIONOF FIELD BINDWEED r, 2, 3, D. C. Britten G. L. Schuster G. J. Michels,Jr. andD. A. owings 3

ABSTRACT

Clippings of Aceria malherbaeNuzzaci infested field bindweed, Convolvulvus arvensisL., wereused to successfullyinfest field bindweedin a greenhousefor up to 70 days after the clippingshad beencollected and storedat either2 and 4"C. Percentageofplants infested was significantly different betweenmites collected from the two sites (34.57" from Moore county and 16.5%from carson county, Texas),and the ability of mites to infest plants decreasedsignificantly as the number of days in storageincreased. No significant difference in percentageof plants infested was found between yearc(27.8o/o in tggg anO 23.2o/orn 2000) or storagetemperature Q3.5% at 2oc and 27.5yoat 4oc). overwinteringl. malherbaefrom field bindweedrootstock infested 66.7 and75%o of the field bindweedgrown in a greenhouse,and it took an averageof 24.5(+ 6.6) and,2l.4(J 4.5) daysfor the bindweed plantsto beginshowing damage symptoms in 1999and 2000, respectively.

INTRODUCTION

Field bindweed,Convolvulus arvensis L., a perennialweed that competeswith other plants for water and nutrients,infests 32 field crops in more than 40 countriei, reducing crop fleld, crop quality, and tand value (Holm et al. 1977). Field bindweed was reported in virginia in 1739, and soon spreadalong the Atlantic seaboard.German and Russian immigrants introducedfield bindweedto the Great Plains as a contaminantin wheat seedin the 1870s(Holm et al. 1977). Field bindweedinfests almost 250,000ha of farmlandin Texasand causeseconomic losses of more than $50 million each year (Bean and Wiese 1990). It canreduce yields of broadcast-seededcereal crops by 20-50%and of row cropsby 50-80%. The annualcost attributedto field bindweedcontrol and its impacton cropsin the TexasPanhandle is $10-60per ha (Beanand Wiese 1990). Field bindweedis foundin dry to moderatelymoist soilsand can survive long periods ofdrought, growingbest in fertile but alsorocky soils (Holm et al. 1977). It is besiadapted to semiaridregions west of the MississippiRiver. The taprootof field bindweedcan a reach depth of 30 m or more, and laterally spreadingroots and rhizomesmake control difficult (swan and chancellor 1976). The seedsare extremelyhard, imperviousto water,and can remain viable in the soil for more than 30 years (weeler and Degennaro l9g4).

' USDA-ARS SouthemPlains Area Conservation and Production Research Laboratory,2300 ExperimentStation Road, Bushland, '^ Texas79012. west TexasA&M university, Division of Agricultwe, p. o. Box 6099g,canyon, Texas 79106. ' Agricultwal J:IT Experimentstation, 2301 ExperimentStation Road, Bushland, Texas 79012. 273 Becauseof their small size and hard coat, the seedsare spreadby birds or wind or as a contaminantin crop seedor manure. Manurecontaining field bindweedseed is a problem in areassuch as the TexasHigh Plains becausefeedlots sell manureto farmersas a fertilizer. Aceria malherDaeis one of eight speciesof eriophyiidmites, and one of four.species in the genusAceria, that attack field bindweedin Greece,Italy, Spain, and southernFrance (Rossenthal1983). A. malherbaewas importedfrom Greeceas a biologicalcontrol agent of field bindweed(Boldt and Sobhian1993). Quarantineof l. malherbaeat the USDA-ARS InsectQuarantine Facility at Temple,Texas, in 1988involved testing for host preference. The mites were testedon plants of the family Convolvulaceae,to which field bindweed belongs,including sweet potato (Ipomoeabatatas (L.) Lam.), ivy-leaf morning-glory(/. hederacaeJacq), tall moming-glory(1. purpurea (L.) Roth),heavenly-blue moming-glory [L tricolor Car.,referred to asI. rubrocaeruleaHook. in Boldt and Sobhian(1993)1, and hedge falsebindweed (Calystegia sepium (L.) R. Br.). No evidenceof feedingwas found on any plant tested. Therefore,Boldt and Sobhian(1993) concluded that A. rnalherbaewas closely linked to the physiologyofthe hostand had a restrictedhost range. A. malherbaefeeds along the centerof the upper surfaceof field bindweedleaves, causingthe leavesto fold and fuse along the middle vein (Rossenthal1983). As feeding progresses,the plant cells hypertrophy,causing the leavesto thicken and developa rough surface.Clear protuberances or papillae,termed microgalls by Boldt andSobhian (1993), are producedby the plant cells and causethe leavesto seemgrainy or mealy. Thesegalls can also occur on the stems,rhizomes, buds, and deeperroots. Heavily-infestedplants are coveredwith microgalls,becoming deformed and chlorotic. The mitesretard the growth of and eventually kill the bindweedplant. l. malherbaealso interfereswith vegetativegrowth, flowering and seedproduction. The mites overwinter on field bindweedroots and attack new rootsand shoots in the spring. Approximately5,800 A. malherbaewere released in May and9,000 were released in Jwre 1989 on field bindweed at the Texas A&M University Agricultural Researchand ExtensionCenter at Bushland,Texas (Boldt andSobhian 1993). The mitessurvived for three consecutiveyears and were consideredestablished in the TexasPanhandle in 1992. The miteswere slow colonizers,moving only 125m per yearfrom the original site. In the fall of 1992,35-38Yoof the leavesof plantssunounding the initial releasesite were infested. In the winter of 1992, plants were dug up and each rhizome was found to be infested with an averageof 78-90mites. It is not known if.A. nalherbaeis dispersedby wind, as are other Eriophyiidae,or distributedon field bindweedby someother mechanism, such as movement from the roots and foliage ofone plant to the next. In 1998,a site whereA. malherbaehad beenreleased east of Amarillo, Texas,was inadvertentlymowed. Threeweeks after the initial releaseat the site, field bindweedinfested with A. malherbaewas found more than 100 m from the releasesite. The mites were thought to have been distributed on clippings from the mowing and had moved off the clippings onto noninfestedfield bindweed. Becauseof theseobservations, we hypothesizedthat a practical and economical way to distribute the mite to new sites might be to store refrigerated clippingsof mite-infestedfield bindweed. Viability of mites on refrigeratedclippings was assessedin this experiment.In addition,since ,{. malherbaeoverwinters on field bindweed roots, we assessedwhether mite-infested rootstock of field bindweedcollected in the fall couldbe usedas an infestationsource.

MATERIALS AND METHODS

Since 1998,the entomologyprogram has beeninvolved in an area-wideproject to redistributeA. malherbaethroughout the Texas Panhandle. Field bindweed sites in Carson and Moore countiesthat were part of this redistributionprogram and heavily infestedwith l. malherbaewere selectedto providemite-infested field bindweedclippings for this research.

274 The field bindweedat eachsite was mowedon 15 October1999 and 25 July 2000, and the clippings were collected in a bag attachedto the mower. Clippings were taken to a gree-nhouse,and 250 ml of clippingswere placed in 946-mlpaper bags. Approximately100 6ags from each site were stored at2 and4oC in Percival I30-BLL environmentalchambers Oircival. Boone. Iowa). Four bags were removed weekly from each environmental chamber, and clippings from an individual bag were distributed over a mite-free field bindweedplant growingin a 15.2-cmdiameter plastic pot. Field bindweedplants, which had been germinated in the greenhouse,were at the four-leaf stage when infested. Clippings were placed on plants on the sameday the clippings were collected from the field (Day 0) and on new plants every 5-7 days thereafteruntil I I setsof plants had beeninfested in 1999 and l0 sets of plants had been infestedin 2000. The plants were observedvisually for damage(microgalls, leaf distortionand chlorosis)caused by mites every 3-5 daysfor three months, beginning three days after the plants were infested. Plants showing damage symptomswere consideredinfested and countedas such. The data were analyzedusing the Probit Procedureavailable in the SAS (1999) computer program and the results of the 12 tests associatedwith this procedurewere usedto determinesignificant differencesbetween field sites,yeaxs, temperatures and lengthsof time mites were in storage. Root sampleswere taken from mite-infestedfield bindweedsites in the fall of 1999 and 2000 to determinewhether mite-infestedrootstock of field bindweedcould be usedas a source for infesting field bindweed with I malherbae. Three to four roots were taken approximately 1.2-m deepfrom the soil. The roots were exarninedfor mites; approximately 50 mites were taken from infestedrootstock and placedonto a field bindweedplant growing in a 15.2-cm plastic pot. Four plants were infested in this mannerat three different dates, giving a total of 12 replications.The foliage was examinedevery 3-5 days for signs of mite damage,beginning three days after the plants were infested. The experiment ended after threemonths. Analysesconsisted of conventionalaverages and standarderror calculations.

RESULTSAND DISCUSSION

Average percentagesof field bindweed plants infested varied by the different combinations of sites, years, and temperatures(Table l). By the end of the experiment, l0- 50% of the plantswere infestedwhen exposedto clippingsof field bindweedinfested with l. malherbae. Variability in percentageof plants infested could have been causedby differencesin the numberof A. malherbaeon the clippings of field bindweedand perhapsthe quality of the clippings themselves. Mowing field bindweed at the nurserieswas, uneven becauseofrough groundand variableplant height. The numberof A. malherbaecollected probably dependedon the percentageof ground coveredand the abundanceof mite-infested field bindweedplants. We did not count individual l. malherbae. Countingeriophyiid mites is labor and time intensive, and the purposeof the experimentswas to determinewhether clippings of field bindweedinfested with A. malherDaecould be storedfor future infestation offield bindweedplants. Chi-squaretests, used to determinesignificant differencesbetween the percentagesof plants infested by A. malherbaecollected from the two sites during the two years and stored at different temperaturesindicated that there were becauseof rough ground and variable plant height. The number of A. malherbae collected no significant differences by year or storagetemperature in the ability of A. malherbae to infest field bindweed (faUte Z). There was a significant difference betweenlocations 1f = tO.Sl, p : 0.05) and the lengthof time thatA. malherbaewere stored (t' = 27.63,p = 0.05). Time in storage results form a gradient based on number of days and a binary assessmentof infestation (l=infested, 0:noninfested), rather than a specific comparisonbetween locations, years, or temperatures. Therefore, only the relevant test results for time in storage are presentedin Table 2, and more detailedresults are presentedin Table 3, The negative

275 TABLE I ' Percentage Infestationof GreenhotseFieh Bindweed Inoculatedwith Aceria malherbae-InfestedBindweed clippng from Two Locationsin the Te>

TABLE2. 12 TestsBetween locatiorl Year,Storage Tenperahre, and Tinp in Storage. Percerfrags SEfor Prob.> " 2 Conparison inGstation SE Btirnat€ X' x2 x. lncatbn Moore 34.5 6.7 1.05 16.57 0.26 0.0001 Carson 16.5 5.9 Year 1999 27.8 6.6 -0.07 0.07 0.25 0.7848 2000 23.2 6.0 TenperanrefC) 2 23.5 6.1 -0.05 0.04 0.25 0.8325 4 27.5 6.5

TirF na -0.03 27.63 0.01 0.0001 " Standarderror. estimate,or coefficient,indicates that successfulinfestation decreased as time in storage increased. The site in CarsonCounty was infestedin 1998,while the site in Moore County was infestedin early 1999. Clippings from the site in Moore County resultedin an overall infestationof 16.50/oof the field bindweedplants in the greenhousecompared to 34.5%with clippingsfrom the sitein CarsonCounty. A. malherbaewas establishedat the site in Carson County almost a year longer than at the site in Moore County and would be expectedto

276 TABLE3. Daysin RefiigeratedStorage and Averap Daysto InfestFbld Birdweedby Aceria malherbaecolbcted fomTwo sites intbe Te4q lg4hardb in 1999ad 2000. 1999,20C lggg,40C 2}oo,2oc 2000,40c Avg days Avg d4a Avg daP Avg dap ' b to infest SE to infest SE to inftst SE to inftst SE MooreCourry .+ic --- 4 8.5 1.5 3 15.3 2.8 3 --- 7 11.0 1.0 7 11.0 0.6 8 15.0 0.0 8 --- t4 33.0 0.0 t4 26.0 3.9 14 t4 21.7 2.3 19 28.0 0.0 19 12.8 1.8 23 17.0 0.0 23 11.5 5.5 22 25.0 0.0 28 15.0 5n 28 2r.0 0.0 35 35 12.0 0.0 34 13.0 1.0 34 10.5 4.s 47 6.0 0.0 47 5.0 1,0 36 15.7 5.0 36 19.3 2.3 50 9.5 1.5 50 I 1.0 0,0 40 16.5 1.5 40 54 7.0 0.0 54 6.3 0.7 49 21.0 0.0 49 21.0 0.0 57 57 12.0 s8 8.0 4.0 58 59 10.0 0.0 59 :: 70 12.0 0.0 70 16.0 0.0 AvB 16.2 I1.6 r4.8 17.3 SE 3.8 2.0 l.l 1.8 CarsonCourny 3 13.0 7.0 3 --- 4 --- 4 26.0 0.0 6 6 23.0 9.0 I I 17.0 ll 19.0 0.0 9 12.0 2.0 9 14.0 0.0 15 15.0 l5 12.5 2.5 l6 ':- l6 23 23 22.0 0.0 23 ).u ;; 23 27.5 ).) 31 3l 32.0 0.0 28 22.0 6.0 28 18.5 9.5 37 37 3l 25.0 0.0 3l 45 r4.0 45 14.0 0.0 44 44 53 53 56 6.0 0.0 56 63 63 59 59 66 66 Av& ;; 20.8 15.3 20.9 SE J.J 2.9 0.9 3.0 u Dap mite-inftstedcfupings were retigerated before beingapplbd to noninftstedfeh birdweed b Stardardenor. 'Mites fiibd to beconp establbhed,rnt inchdedin averap'

result in greaterinfestation of the field bindweedplants. However, the CarsonCounty site wzrsmowed in October, and the Moore County site was mowed in July. At the Carson County site, A. malherbaemay have either left the plant or moved down onto the roots by October,but may have still beenon the foliage at the Moore County site in July. We originally hypothesizedthat the longer l. malherbae-infestedclippings were stored,the less likely they were to infest field bindweedbecause of declining clipping 277 freshness, of saprophytic fungi and A. marherbae g:neral, .erowth moving off the cliippings. In thiswas true, although the number of daysit tookfor tf,. no"iriJiiTffi*."; plantsto becomeinfested by the mite-infested clippingswas not r,igrrry"*iire fiable 3). Regardlessof the site,year, or temperature,rewer fieldbindweeO ptan:ts were inresteO uy,a. malherbaeas the number of days that the mites were in storageincreased (Fig. 1, 2). Percentageof plants infestedafter being exposedto clippings f.om"the rit" in voor. County (Fig' l) was less than that from exposureto clippingi collectedfrom the site in Carson county (Fig. 2). At most, 50%oof field bindweed'plints in m" gr"";o;re were infested when exposedto A. malherbaeon refrigeratedclippings from Moolrecouniy wrril" as many as 75%oof plants were infested when clippings-from the site in carson county were refrigeratedfor as many as 54 days.again, ihls could be becauseof the time of year the clippingswere taken.

100

o r9992c. '7< E 19994C d a 2OO02C o 2000 4c I ----OverallAverase I j--- _-'---?- | A

Days in Storage FIG. l. Percentageof field bindweedplants infested withl. malherbaeafter exposureto mite-infestedclippings of field bindweedfrom Moore county, Texas,in 1999 and 2000 versusnumber of daysthe clippingswere stored at2 or 4C.

In 1999,A malherbaefrom clippingsfrom the site in Moore Countyremained viable for at least59 and57 dayswhen the clippingswere stored at2 and4oc, rispectively(Table 3). In 2000,A. malherbaefrom clippingsfrom this site remainedviable for at least7d days whenthe clippingswere stored at 2 and4oc. In 1999,A. malherbaefrom infestedclippings storedat 2oCfrom the sitein Moore Countytook zrnaverage of 16.2a3.g daysto infeif neiA bindweedand an averageof 11.6+ 2.0 daysto infestfield bindweedw-hen the clippingswere storedat 4oc. In 2000,A. malherbaefrom infestedclippings from the site in Moore-county that were storedat 2oC took an averageof 14.8+ l.l daysto infestfield bindweed,and an averageof 17.3+ 1.8days to infestfield bindweedwhen stored at 4oC. In 1999,clippings from CarsonCounty harbored live mitesfor 56 and28 dayswhen storedat 2 and4"c, respectively(Table 3). In 2000,A. malherbaeremained viable for 45 days when the infestedclippings from this site were stored at 2 or 4"c. ln 1999,A. malherbaefrom infestedclippings storedat 2oC from the site in CarsonCounty took an averageof 13.8 + 3.3 daysto infest field bindweed,and an averageof20.8 + 2.9 daysto infest field bindweedwhen the clippingswere storedat 4"c. In 2000,A. malherbaefrom infested clippingsthat werestored at 2oc from carsoncounty took an averageof 15.3t 0.9

278 q AI\ ,o

t50

o Pr{ a-'

10 20 30 50 60 70 80 90 r00 Dap in Storage FIG. 2. Percentageof field bindweedplants infested with Aceria malherbaeafter exposure to mite-infestedbindweed clippings from CarsonCounty, Texas, in 1999and 2000 versus numberof daysclippings were stored at2 or 4oC.

days to infest field bindweed, and an averageof 20.9 t 3.0 days to infest field bindweed whenthe clippingswere stored at 4"C. The experimentsin 1999and 2000 to determineif overwinteringmites collectedfrom field bindweed rootstock could infest greenhouse-grownfield bindweed resulted in infestationrates of 66.7 ard75.0%oand averages of 24.5S 6.6) and21.4 (!4.5) daysfor the bindweedplants to begin showingdamage symptoms, respectively. The results of this research indicated that mowing A. malherbae-infestedfield bindweedand using the clippings as a sourcefor infestationis a practical way to redistribute this mite as a biological control agent. If mite-infestedfield bindweedclippings are storedat either 2 or 4oC and used in a redistributionprogram, the infestationrates observed would be acceptable, especially if the clippings were used within one month. However, for commercial application the infestation rate is low, and additional quality control would be neededto ensure greater rates of infestation. Low rates of infestation after exposureto mite-infestedclippings are probably related to variations in the abundanceof mites, suitability of mowed plants to harbor l. malherbae,and variations in the efficiency of the mower. Greaterinfestation rates during longer periods in storagecould probably be obtained if A. malherbaewere uniformly distributedin fields from which the field bindweedclippings werecollected and the bindweedwas activelygrowing. Mower bladetype, blade sharpness, and mowing height also needto be considered.The ability to useoverwintering mites from field bindweed rootstock as an infestation sowce is of value becausemites could theoreticallybe collectedat anytime of the yearfrom rootsto provideA. malherbaecolonies for usein a greenhouseor laboratory.

ACKNOWLEDGMENTS

We thank JohnnyBible, ShanaCamarata, Lana Castleberry,Jeremy Holmes (Texas Agricultural ExperimentStation, Bushland) for assistingwith collectionsand processing samples.We alsothank Arden Colette(West Texas A&M University)for assistingwith the statistical analyses,and Mark Lazar (TexasAgricultural ExperimentStation) and Brent Bean L 279 (Texas CooperativeExtension Service) for their reviews of the manuscript. This research was part of a thesis submittedby the senior author in partial fulfillment of the requirements for a Master's degree in Agriculture from west Texas A&M university. Funding was provided by the Texas Agricultural Experiment Station, with partial support through,grant IPM01-014from the TexasDepartrnent of Agricultwe.

LITERATURE CITED

Bean,B. w., andA. F. wiese. 1990. Field bindweedcontrol in the TexasHigh plains.Tex. Agric. Ext. Serv.Publ. L-2339. Boldt, P. E., and R. Sobhian. 1993. Releaseand establishmentof Aceria malherbaeAcarj: Eriophyiidaefor conffolof field bindweedin Texas.Environ. Entomol. 22:234-237. Holm, L. G., D. L. Plucknett,J. V. Pacho,and J. P. Herberger. 1977. The World's Worst Weeds,Distribution and Biology. University Press of Hawaii,Honolulu. Rossenthal,S.S.1983.Cunentstatusandpotentialforbiologicalcontroloffieldbindweed withAceria convoltuli, pp. 57-60. InM. A. Hoy, L. Knutson,and G. L. Cunningham [eds.]Biological Control of Pestsby Mites.Univ. Calif. Agric. Exp. Stn.publ. 3304. SAS Institute.1999. SAS Procedures Guide, Version 8. SASInstitute, Cary, NC. Swan,D. G., and R. J. Chancellor. 1976. Regenerativecapacity of field bindweedroots. WeedSci. 24: 306-308. Weeler,S. C., and E. P. Degennaro.1984. Growthand reproductive characteristics of field bindweed(Convolulus arvenis) biotypes. Weed Sci. 32:525-528.

280 vol.28 NO.4 SOUTHWESTERNENTOMOLOGIST DEC.2003

NUCOTN 33B" A].ID DBLTA A}iID PINELA}iID COTTONS: PINK BOLLWORM (LEPIDOPTERA: GELECHIIDAE) INFESTATIONS AllD CRYIAC TO)flC PROTEIN IN OVERWINTERED AI{D SEEDEDCOTTONS WTTII BIOASSAY MORTALITIES OF OTHER LEPIDOPTEROUSLARVAE.

T. J. Henneberry,L. Forlow Jecll and T. de la Torre USDA, ARS, PWA, WesternCottonResearch Laboratory 4135E. BroadwayRoad Phoeni:qAZ 85040-8803

ABSTRACT

NuCOTN 33Bo (BO and Delta and Pineland (DPL 5415) cottons were grown in furrow and furrow plus drip inigated cotton plots in Arizona. CrylAc toxic protein produced in Bt cotton was measured with commercially available enzymeJinked immunosorbentassay @LISA) kits. Irrigation type had no effect on CrylAc measuredin cotton leavesor bolls. First-instar,laboratory-reared pink bollworm (PB!V), Pectinophora gossypiella (Saunders),larvae placed on Bt cotton bolls collected in field throughout the seasondied (10070)irrespec'tive of decreasingCrylAc toxic protein lwels in late season bolls. PBW field infestationsaveraged over 2.5 live larvaeper (DPL 5415) immature green boll compard with no live larvae in Bt immaturegreen bolls. Resultswere similar for PBW infestations of open mature bolls on Bt and DPL 5415 whole cotton plant samples.Two unexplain€dPBW lanal exit holesocruned in openmature Bt bolls. These could have been on plants from non-Bt contaminatedseed in Bt seedlots. In bioassayg using laboratory-rearedcabbage looper (CL), Triclwplusia ni (tliibner), larval mortality percelrtagesdeoreased with decreasingCrylAc contentmeasured in Bt leaves. CrylAc protein was expressedin cotton bolls and leavesfrom overwinteredBt plants. PBW and tobacco budworm (TBW), Heliothis irescens (L.), larval mortatities averaged99 and 100% (overwinteredand 2002 seededBt cotton), respectively,compared with averagesof 40 and 3olo,respectively, on DPL 5415 bolls and leaves(overwintered and 2002 seeded cotton).

INTRODUCTION

Transgeniccottons (80 with the Bollgardogeneproducing the insecticidalprotein of Bacillus tlruringiensis Kurstaki @erlinger) have been widely grown commercially in Arizona since 1997 (Sims a il. 2@2). Pink bollworm (PBW), Pectinophoragossypiella (Saunders),larvae have beenvery susceptibleto the toxic protein in the field. Annual widespreadand intensive monitoring has shown

METHODS A}.ID MATERIALS

2001 Field Plors. DPL 5415 and NuCOTN 33Bo @t) (MonsantoCo., St. Louis, Mo) cotton seedswere planted on 13 April at the western cotton ResearchLaboratory, Phoenix,AZ. There were four whole plots of eachcultivar. Eachwhole plot was 16 rows wide by l9-m long. Split plots were 8 rows wide by l9-m long. Split-plot treatments (two) were (l) funow irrigationsat l4-day intervalsthat suppliedabout l2.7cm of water and (2) furrow inigation plus supplementaryT-tape (T-systemInternational, Inc., San Diego, CA) drip irrigation. Supplementarydrip irrigations were 102 liters of water per hour per day per 3lm of row. The orperiment was replicatedfour times. The T-tape drip system was controlled with a timer-operated solenoid valve that was set to apply supplementarywater to furrow irrigated plots for I h durationson eachday. Male PBW moth catches. PBW male mothswere sampledby placing a Delta trap (Flint and Merkle 1983),baited with one gossyplure(Hummel €t al. 1973)impregnated rubberstopper (100 Fl), in eachquadrant ofthe field. Trapswere locatedat the tops of cotton plant canopy. They were collected and replacedweekly at canopy level with new traps. Glossyplure-impregnatedrubber stoppGrs{*Gr€ replaced every other week. Laboratory bimssays. PBW and CL lanrae used were from the Westem Cotton ResearchLaboratory colonies reared on artificial diets describedby Bartlett and Wolf (1985)and Henneberry and Kishaba (1966), respectively. For CL bioassaysin 2001, Bt and DPL 5415 leaveswere picked at randomfrom eachplot on 19 June,l0 July, 14 Augustand 25 September.Leaves were trimmed to fit in l5.0cm diameterx l.5cm deepplastic petri disheslined on the bottom with moist filter paper. Five first-instarCL larvaewere placedin eachdish containinga Bt or DPL 5415 leaf. Living and deadlarvae were countedafter serrendays. On ll July, 15 Augustand 18 September2001, firm, immaturegreen Bt andDPL 282 5415bolls (21to 28 days-old)were picked ftom plantsin eachplot. Bolls (5) wereplaced individually in 5cm diameterx 6.5cm tall cylindrical polyahylene containers. Fivg first- instar PBW larvae were placed on each boll. Bolls were examinedon day 7 following infestation with the aid of a microscope, PBW larval €ntranc€holes in the carpel.walls were counted. Bolls wer€ then dissected, and all living and dead larvae and their dwelopmental stageswere recorded. F,LISA. The amounts of CrylAc protein in Bt cotton leaves and bolls were determinedon all sampling dates using enzrym+linkedimmunosorbent assays @LISA). Controls were DPL 5415 leavesand immanre cdon bolls. Tissre sampleswere taken fiom the same plant parts used for lanral infestations. Materialg sample preparations, solutions, odractionq dilutions, negative controls, and assayswere as describedin the CrylAb/CrylAc plate kit @nvirologi:q Inc., Portland, ME). Sampleswere 0.6cm boll piecesthat were weighed and placed in l.5cm microcentrifirgetubes. Leaf samples,also weighed, w€re 0.6cm diameter, ciroular leaf punches. Sampleswere homogenizedin o

RESULTS

2001Male PBW moth catches. Moths caughtin glossyplure-baitedtraps increased in August, decreasedslightly in early September,and peak numbers occurred in early October, followed by a leveling off and sharp decreasein late November (Fig. lA). The data confirm the presenceof PBW adult season-longactivity in the field during boll developmentand maturation. Laboratory bimssays. There were no significant effects of furrow irrigation comparedwith funow plus drip irrigation for either cultivar for living or deadPBW or CL larvae,percentage mortdities, PBW entranceholes or other insectcriteria (F valuesfor 1,33df in all caseswere < 1.68;P > 0.05). Thusthe datawere combinedfor all further statisticalanalyses of 2001PBW andCL data. CrylAc insecttoxic proteinfound in leaves from furrow irrigatedplots (1.05,0.24,0.40 and 6.18 ppm on l9 June,l0 July, 14 Augrrst and 25 Septembeqrespectively), were not significantly different comparedwith leaves from furrow plus drip irrigatedplots (0.86, 0.074,0.54 and 0.23 ppm, respectively)(F valuesl, 14dFl. 14,0.79, 1.52, 1.23, respectively. P > 0.05) AverageCL larval mortalitieswere 81, 63,32, and6Yo for Bt larvaesampled 19 June, l0 July, 14 August and 25 September,respectively (Table l). Corresponding averageCrylAc toxic proteinamounts (ppm)^measured were 0.96,0.69,0.47and 0.18, respectively. The regression,y3.12+66.6x, (4.99 was significant(P < 0.05, wherex and y are dose-mortalityrelationships). Larval mortatities feeding on DPL 5415 were 3, 13,3 and9/oon19 June,l0 July, 14 Augrst and25 Septernber,respectively. For boll samples taken on ll July, 15 August, 18 September,CrylAc determination(ppm) were 0.13 vs 0.14,0.11vs 0.ll and0.05 vs 0.04ppr4 respectively, for furrow alone comparedwith furrow plus drip inigation (t:0.59, 0.20 and 1.60, respectively,62 df, P>0.05). Inigation tlpe effectson CrylAc toxic proteindetermination in bolls were not significant on any samplingdate. AveragePBW larval mortalities over all dateswere shown in Table 2. No laboratory-rearedPBW larvae survived beyond the first instarafter placement on Bt bolls. On dissectionof DPL 5415bolls, there were 0.03, 0.25 and 0.35 living first-instar, second through fourth instar and pupae per boll,

284 A. 2001

lr .gf = lz eT' F \0 o t +a Oa B. 2002 = -!P Eo- C' z a

30102030 9 192t 9 rt2l 8 18'6 f i n I i 2f 7 1f 27 a M.y Jun Jd Aug sep Oct ibv t'.c

FIG I . Mean(+ SE)numbers of pinkbollworm male moths caught per trap per night in split-plotplantings ofNuCOTN33B- and Deltapine 5415 cottons.

Table l. Mean (+ SE) Percentagesof Laboratory-RearedCabbage Looper (CL) Larval Mortality Following SevenDay FeedingPeriods on NuCOTN 33Bo(Bt) andDPL 5415 LeavesFrom Field Plots. No." Bt CrylAo DPL5415 Samplingdate Samples % Ivfiortaliw (ppm) % Mortalitv June19 8 8l.ll + 4.55 0.96+ 0.09 2.52+ 1.30 July l0 16 62.50+ 8.43 0.69+ 0.06 13.24+2.18 August 14 16 32.23+ 5.28 0.47+ 0.05 2.92+ t.l2 September25 16 5.59+ 0.18 0.18+ 0.04 9.41* 2.94 ' Means of 4 replications, 2 to 4 observationsper replication. Five CL larvae per observation.

285 ct I 9so eg o \ o € E \o tr E= 5 (t c, 66 q Ou;

c't o

= \o t c\ o c I -,i cl d

a q = o o ()tr .c)

@ = tt F o tr o o qt t c o o ci I .9p a b g a o 0) tatct lcl T c't .E c- tb!) F IB t'? I'| ,a; l€A VI

q, o +r^ \OO \OO

E,tr I !E E 00 0 c o o ts 6 b rn9 C) 6 N a *6 t) q, Fl E E +-i t o e o 6l E Eia, E tr 11

286 respectively. The averageCrylAc content for the three boll collection dateswas 0.10 + 0.01 ppm (not tabulated). CrylAc amountsranged from 0.148 ppm in early-seasontg 0.024ppm in late-season Field Infestations. Immature cotton bolls picked at random from each plot in the field on 28 Septemberand 24 Octoberhad no living PBW larvaein Bt bolls and2.06 dead lawae per boll comparedwith no deadlarvae in DPL 5415bolls and2.63living larvaeper boll (Table 3). There was no significant difterence betweencultivars for the number of PBW entranceholes per boll. Individual tarvae may make more than one entrancehole before enteringa cotton boll (tlenneberry and Jech2000).

TABLE 3. Mean (+ SE) Numbers of Pink Bollworm (PBIV) Lanae per Field Infested Immafire CrreenNuCOTN33Bo or DPL 5415Conon Boll Sampledon 28 Septemberand 14 October2001. Iarvae ger boll' Cultivars No. bolls Live Dead EntranceHoles NuCOnI33B'(BD 400 0.00+ 0.00b 2.06+ 0.53a 5.25+1.21t DPL 5415 zl00 2.63+0.71a 0.00+ 0.00b 8.88+ 1.84a F, l,2l df 13.26 49.3 3.n P s 0.05 < 0.05 > 0.05 ' Means of 4 replications,2 observationsper reptication in a column not followed by the same letter are significant by different P < 0.05. Metlrcd of least significam differences. Sampled28 Septernberand 24 O6obq,200l.

For whole plant samplesoverall, there were more open bolls per ootton plant in furrow plus drip inigated comparedwith furrow irrigated plots (table 4). More PBW

TABLE 4. Mean (+SE) Numbersof Open Cotton Bolls per Plant per Node on Whole Plant SamplesofDeltapine 5415 and NuCOTN 33Bo Cottons. Numberof PBWLarvae InfestedBolls CfrltivarVlrrig*ion Openbolls PerBoll f/o)' NuCOTNB33*@tf Furrow 3.53+ 0.35a 0.01+ 0.01b 0.17* 0.17b Furrow plus drip 3.86+ 0.33a 0.03+ 0.02b 0.46+ 0.34b Deltapine5415 Furrow 4.23+Q.42 a 0.04+ 0.02b 1.36+ 0.79b Funow plus drip 3.17+ 0.41a 0.46+ 0.12 a I l.6l + 3.49a 4 (P), df: l, e5 2.46(0.05) 14.79(P 50.0s) r r.rs (s o.os; ldain effects Cnltivar Btb 3.70+0.23t 0.02+ 0.01b 0.32+ 0.19b DPL 54T5 3.70+ 0.30a 0.25+ 0.07a 6.48* 1.94a 4 (P),df= 1,95 0.00e 0.05) 20.88(S 0.05) 1e.63(s0.0s) Irrigation Furrow 3.35*0.27 b 0.03+ 0.01b 0.75+ 0.40b Furrow plus drip 4.04+Q.27a O.2A+0.07a 6.03+ 1.90a F (P) df = l, e5 8.78(50.05) 17.70(s 0.05) 13.94(S o.o5) ' Meansof 4 replicationswithin the samecolumn and category not followd by the same letterare significantly different P < 0.05.Method of leastsignificant differences. o No orit holes, all dead larvae.

287 larvae per boll occuned in fuirow plus drip irrigated plots comparedwith furrow only inigated DPL 5415, but not Bt plots. There was an av€rageof 0.25 pBW larvaeper DpL 5415 boll comparedwith 0.02 PBW larvaeper Bt boll, The few pBW found in Bt bolls occurredon ftuiting branchesat nodes6 to l0 and 1l to 15 (Table5). No pBWlarvae were found in DPL 5415 cotton bolls from fruiting brancheson nodes6 to 10, but generallythereafter, larvae per boll increasedwith increasingnode numberexcept for late in the seasonwhen larvae per boll decreased(nodes 26-35). The highest percentagesof DPL 5415bolls infestedoccurred for bolls on fruiting branchesat nodes3l to 35. Datais not tabulated,but no PBW infestationsoccurrd in Bt or DPL 5415 immature greenbolls from fruiting branch nodes6 to 25 shownin Table 5. No infestedimmature green Bt bolls were found at nodes26 to 30 and3l to 35 comparedto 5.6 and40.0% infestation for DPL 5415immature green bolls from the samenodes.

TABLE 5. Mean (* SE) Numbersof l\datureOpen Cotton Bolls and Pink Bollworm Larvae prr B"tt "t Dttr".""t Pt"t N.d". Numberof Cuhivar/ OpenbollJ Bolls Infested PlantNodes = Node Series PBW larvae/boll (7o) NuCOTNB33*@t) Nodes 06-10 4.08+ 0.28 0.05+ 0.03b-d 0.96* 0.63cd r l-15 4.80+ 0.37 0.05+ 0.05cd 0.08+ 0.93cd t6-20 3.80+ 0.35 0.00+ 0.05d 0.00+ 00.00d 2t-25 5.00+ 0.56 0.00+ 0.00d [email protected] 2630 3.38+ 0.33 0.00+ 0.00d 0.00+ 00.00d 3l-35 L13+0.26 0.00+ 0.00d 0.00+ 00.00d DPL4515 Nodes 0Gl0 3.98+ 0.29 0.00+ 0.00d 0.00+ 0.00d I l-15 5.05+ 0.40 0.05+ 0.05cd 0.74+ 0.73cA t6-20 4.to+0.67 0.25+0.t2&, 4.69* 2.36bc 2l-25 5.58+ 0.64 0.65+ 0.26a I1.55+ 4.07a 26-30 2.88+ 0.47 0.30+ 0.18b 7.22t 4.08a-c- 31.35 0.60+ 0.29 0.25+ 0.18 bc 20.67+ I I .00ab F, df= 5,69 0.62 3.98 3.43 P > 0.05 < 0.05 < 0.05 oMeans of 4 replicationg5 whole plantsp€r r€plicationin a columnnot followed by the same letter are sigrificantly different. Method of leastsignificant differences.

Cotton Yields. Numbers of cotton bolls and cotton seed and lint weights were numbrically but not statistically greaterfor Bt comparedwith DPL 5415 cotton (Table 6). Numbers of cotton bolls and higher cotton seedand cotton lint weights were significantly greaterfor firrrow plus irrigated plots comparedwith furrow irrigation. 2002 Male PBW Moth Catches. In20O2, fewer PBW male moths were caught in gossyplurebaitedtraps comparedwith 2001 @ig. lB). Lessthan one moth per trap per night was caught through mid-Septemberfollowed by a rapid increaseto a peak of five moths per trap per night in late October and a suddendecline to about three tenths of a moth per trap per night in early November. Laboratory Bioaswys. No live TBW larvae survived following three-dayfeeding periods on leavespicked from overwinteredor 2OO2seeded Bt plants (Table 7) compared

288 TABLE 6. Mean (+ SE) Numbersof OpenMature Cotton Bolls and Cotton Seedand Lint Weights per Four Maers of DPL 5415 and NuCOTN 33Bo Cotton Cultivar in Funow and F ro* Ptu$Dttp h"gatto" Ptots Cramsof cotton(per 4 m of row)' No. bollsf CultivarVlrrication 4 mrow S€€d Lht NUCOTNB33"(Bt) Furrow 624+68 a ll90 + 105a 7ll +61 a Furrow plus drip 760+34a l57l + 87a 975+54 a DPL 5415 Furrow 454*72a 817+13a 547+90 a Furrow plus drip 72O+46a l44l + 89a 9El+62a FO(P) l.o7 (> 0.0s) 1.05(> 0.05) 1.38(> 0.05) tvlain effects Cultivars 692*44 a 1380+ 96 a 843+63 a Bt DPL 5415 586+64a ll28+ l39a 769*98 a F"(P) 2.81e 0,0s) 4.43(> 0.05) o.e3(> 0.05) Inigation Furrow 539+56b 1003+ 105b 629+59b Furrow plus drip 74O+28a 1506+ 630a 983+38a F " (t\ 10.3l (s 0.05) 17.68(s 0.0$ 21.3(s 0.0s) " x l0O0 approximateper acre equivalentgmeans of 4 replicationswithin a column in the samecategory not followed by the sameletter are significantlydifferent (P S 0.05). Method of leas significantdifferences. o df- 3, 9; P = probability. 'df: 1,9.

TABLE 7. Mean Numbers of Laboratory-RearedLMng and Dead Tobacco Budworm Larvae and PercentageMortalities aner fhreeOay FeedingP€riodson NuCOTN 33Bo (Bt) a"d Dettaptne(DPL) 5415Cott*I*"tet fro. r00l Cultivar/ TBWkrvae CrylAcToxic Treatment' Ptotein(PPMto CultivarEffecl Bt 0.00b 3.98a 100.00a 0.379- DPL 4.13a 0.10b 2.40b 'T- F=" r974.20 t423.30 2386.l0 P: 0.000 0.000 0.000 2002 Seededvs VolunteerEffect Seeded 2.lOa 2.O2a 51.98a 0.416a Volunteer 2.02 t 2.06a 5Q.42a 0.342a F= 0.6 0.01 1.43 p= > 0.05 > 0.05 > 0.05 'Means of 4 replications,3 obs€n/ationsper replicationin a column in the samegroup not followed by the sameletter are significantlydiflerent. bt = 0.502,22 dq NS. " df: l, 33.

289 with over 97%TBV.I larval survival following the samefeeding period on overwinteredor 2002 seededDPL 5415 leaves. Amountsof CrylAc toxic protein found in leavesfrom overwintered Bt plants was not significantly different comparedwith amounts in leaves from 2002 seededplants. For PBW lanrag more dead and fewer live larvae were found in Bt bolls from overwinteredor 2OO2seeded cotton plantscompared with bolls from overwinterd or 2002 seededDPL 5415plots (Table 8). The differencesin CrylAc toxic proteincontent in bolls of overwintered volunteer Bt cotton compared with 2002 seeded cotton were not statisticallysignifi cant.

TABLE 8. Mean Numbersof Laboratory-RearedLiving and Dead Pink Bollworm Larvae and PercertageMortalities After Seven-DayFeeding Periods in NuCOTN 33Bo (Bt) ild 5415CottonBolls from 2001Overwintered and20f2 SeededCottons. Cultivar/ PBWLarvae Ac Toxic Treatment" %Mortalily Prot€in(PPMb CultivarEffecr Bt 0.01b 1.57t 99.81a DPL 1.00a 0.65b 40.20b F= 6t.72 t9.t7 63.71 P= < 0.05 s 0.05 < 0.05 2002 Seededvs VolunteerEffecl Seeded 0.49 a 1.23t 71.18a 0.332t Volunteer 0.52a 0.99a 68.83a 0.307a F= 0.05 1.29 0.08 P= > 0.05 > 0.05 > 0.05 'Mean of 4 replicationg 2 obsenationsper replicationin a column not followed by the same letter are significamlydifferent. " t = 0.309,14 d[ NS

DISCUSSION

Bt cottonscontirme to maintain high levels of efficacy for seasonlong PBW control in Arizona(Sims et aI.2002). The author'sstatewide monitoring of PBW susceptibilityto CrylAc since 1997 shows higher levels of susceptibilityto Bt toxic protein in 2000 comparedwith 1997. In our sh.rdies,CrylAc amountsin leaf and fruiting body tissuesas determinedwith commercialELISA kits decreasedas the seasonprogressed. PBW larval mortdity in all field sampledbolls was lWo at the lowest level of CrylAc measured. Two PBW exit holes, found in mature openbolls, were probably from plants from non-Bt seedthat may occur as contaminantsin Bt seedlots (Flint and Parks 1999). CrylAc toxic protein amountsin overwinteredBt cotton leavesand fruiting forms were not significantly different comparedwith amountsmeasured in cotton leaf and boll tissue from cotton seed from the current year. Low toxic protein extraction effrciency and interferenceof other chemicalsnot showinginsecticidal activity may haveinfluenced our quantificationsof the toxic protein (Greenplate 1999) comparisons. However, the insect bioassaysprovide widence that zufficient lwels were expressedby overwinteredBt plantsthat no difference in efficacy againstPBW or TBW larrraeur€fe d€tectedcompared with tissuesfrom 2002 seededcotton. Cotton is a perennial plant and how long the plants continue to express CrylAc protein after sequentialperiods of dormancyand regrowth is unknown' Cabbagelooper lanrae af,eless susceptibleto the Bt toxio protein than PBW or TBW (tlenneberry €t al. 2001b) and larval mortdity perc€ntagesdecreased late in season and appearedto be relatedto lower CrylAc @ntentsin the leef tissue. The resultssuggest 290 a dose responsereaotion for this speciesthat corld resrlt in reduced controt late in the season. Howwer, our results need vcrification since CL mortalities were recorded after seven-day feeding periods. CL mortalities have been daermined to increase with increasinglylonger feedingperiods on Bt cdons (Ienneberry et al. 2001b).

LITERATI,JRECITED

Bartleq A. C. 1995. Resistanceof the pink bollworm to Bt transgeniccottorq pp. 766- 768. In D. J. Richter and J. Armour [eds.] Proc. Beltwide Cotton Prod. Res. Conf , Natl. Cotton Council of Amer., Mernphis,TN. Bartlett A. C., andW. W. Wolf. 1985. Pecfinqhua gossypiella,pp. 415-430. 1n R. F. Moore and P. Singh [eds.]Ilandbook of insectrearing, vol. 2. ElsevierScientific PublishingCo., Amsterdam. Ellswortb P. C., D. Meade ud J. Diehl. 1996. Bt Cotton in Arizona: 1995 Variety Rezults. Univ. of Arizona CooperativeExtensio4 Tucsoq l&, 2 pp. Ellswortb P. C., J. Silvertoot[ and J. W. Diehl. 1995a. Bt cotton in Arizona: 1994 Variety Results. Univ. of Arizona CooperativeE:rtensioq Tucro4 1fl,2 pp. Ellswortb P. C., J. W. Diehl, and T. F. Watson. 1995b. Bt cotton in Arizona: What will change?Univ. of Arizona CooperativeExtension5 Tucsoq AZ, 2 pp. Flint, H. M., and J. R. Merkle. 1983. Methodsfor eflicient use of the deltatrap in the captureof pink bollwolm moths. Southwest.Entomol. 8: 140-144. Flint, H. M., and N. J. Parts. 1999. Seasonalinfestation by pink bollworn\ Pecfinoplnra gossypiella (Saunders), of transgenic and non-transgenic cultivars of cotton, Gossltpiumhirsutum L., in Central Arizona. Southwest.Entomol. 24: 13-20. Flint, H. M., L. Antill4 J. E. Leggett, andN. J. Parks. 1996. Seasonalinfestation by pink bollworm, Pectinoplnra gossypiel/c(Saunders), of transgeniccotton containingthe Bollgard t gene, planted in commercialfields in central Arizona. Southwest. Entomol.2l:229-235. Flint, H. M., T. J. Henneberry,F. D. Wilso4 E. Holguit\ N. Parks, and R. E. Buehler. 1995. The effects of transgeniccotto4 Gossltpiumhirstum L., containingBacillus thwingiensis toxin genes for the control of the pink bollworrq Pectinophora gossypiella(Saunders), and other arthropods. Southwest.Entomol. 20:281-292. Frisvold, G. B., R Tronstad,and J. Mortensen. 2000. Adopion of Bt cotton: Regional differencesin producercosts and rehrrn, pp. 337-240. /z P, Dugger and D. Richter [eds.] Proc. Beltwide Cotton Prod. Res. Conf., Natl. Cotton Council of Amer., Memphis,TN. Gould,F., and B. Tabashnik. 1998. Bt-cottonresistance management, pp 67-105. 1n M. Mellon and J. Rissler [eds.]Now or Never SeriousPlans to Savea Natural Pest Control. Union of ConcernedScientists, Two Brattle Squarg Cambridge,Mass. Greenplatg J. T. 1999. Quantificatlon of Bacillus tlruringiensis insect control protein CrylAc over time in Bollgardocotton fruit andterminals. J. Econ.Entomol. 92: 1377-1383. Greenplate,J. T., G. P. Head,S. R. Pen4 andV. T Kabuye. 1998. Factorsinfluencing the survival of Helica,erry zea on Bollgardo cofror\ pp. lC2C-lO22. /n P. Dugger and D. Nchter [eds.]Proc. BeltrvideCotton Prod. Res. Conf., Natl. CottonCouncil of Amer.,Memphis, TN. Henneberry,T. J., and A N. Kishaba 1966. Cabbagelooperg pp461478. InC.W. Smith [ed] tnsect colonization and mass produotion. Academic Press,Inc., New Yorlq l.IY Henneberry, T. J., and L. Forlow Jech. 2000. Seasonalpink bollwonrq Pectinophora gossypiella (Saunders), infestations of transgenic and non-transgenic cottons. Sonthwest.Entomol. 25: 273-286. 291 Henneberry,T. J., L. Forlow Jec[ andT. de la Tone. 2001a. Effectsof transgeniccotton on mortality and developmentof pink bollworm (Lepidoptera:Gelechiidae)larvae. Southwest.Entomol. 25:115-128. Henneberry,T. J., L. Forlow Jecb and T. M. de la Torre. 2001b. Effectsof transgenic cotton on cabbagelooper, tobacco budworm and beet armyworm (Lepidoptero: Nocuidae) larval mortality and development and foliage consumption in the laboratory. Southwest.Entomol. 26: 325-338. Hummel, H. E., L. K. Gaston,H. H. Shorey,R. S. Kaae, D. J. Byrne, and R. M. Silverstein. 1973. Clarificationof the chemicalstatus of the pink bollworm sex pheromone.Science. 181: 873-875. Marchosky,R. P.,^P.C. Ellsworth,H. Ivfloser,and T. J. Henneberry.2001. Bollgardoand 'DP Bollgud [I" efficacyin new isogeniclines of 50' uplandcotton in fuizona"p. 866. In P. Dugger and S. Richter [eds.] Proc. Beltu,ide Cotton Prod. Res. Conf,, Natl. CottonCouncil of Amer.,Memphis, TN. MSTAT-C. 1989. MSTAT-C, a microcomputerprogram for the desigrl managementand analyses of agronomic research experiments. Michigan State University, East Lansing, M. PatinrA. L., T. J. Dennehy,M.A Simq B. E. TabadrnilqY-B Liu, L. Antill4 D. Gouge,T. J. Henneberry,and R. Staten. 1999. Statusof pink bollworm srsceptibility to Bt in Arizon4 pp. 991-996. ^Iz P. Dugger and D, Richter [eds.] Proc. Belrwide Cotton Prod. Res. Conf, Natl. Cotton Council of Amer. Memphis. TN. Simmons,A. L., T. J. Dennehy,B. E.TabashnihL. Antilla, A. Bartlett,D. Gougg andR. Staten. 198. Evaluation of Bt cotton deploymentstrategies and efficacy against pink bollworm in Arizona"pp. 1025-1030. InP. Dugger and D. Richter [eds.] Proc.Beltnide CottonProd. Res. Conf., Natl. CottonCouncil of Amer.,Memphis. TN. Sims,M. A., T. J. Dennehy,A. Patiq Y. Caridre, Y-B Liu, and B. Tabashnik. 2001. Arizona's multi-agency resistancemanagement program for Bt cotton: Sustaining the susceptibilityof pink bollwonq pp. l175-1179. InP. Duggerand D. Nchter [eds.] Proc. Beltwide Cotton Prod. Res. Conf, Natt. Cotton Council of Am€r., Memphis,TN. Sims,M. A., T. J. Dennehy,L. Shriver,D. Hulley, Y Carridrg andB. Tabashnik. 2OO2. Susceptibility of Arizona pink bollworm to CrylAg CD-ROM. .Iz P, Dugger and D. Richter [eds.]Proc. BeltwideCotton Prod. Res. Conf, Natl. CottonCouncil of Amer.,Memphis, TN. Watso4 T. F. 1995. Impactof transgeniccotton on pink bollworm andother lepidopteron pests,pp 759-760.In P. Duggerand D Richter[eds.] Proc. Beltwide Cotton Prod. Res.Conf., Natl. CottonCouncil of Amer.,Memphis, TN. Wilso4 F. D., H. M. Fling W. R. Deaton,D. A. Fischhotr,F. J. PerlalgT. A. Armstrong, R. L. Fuchs,S. A. Berberict\ N. J. Parks,and B. R. Stapp. 1992. Resistanceof cotton lines containing a Bacillus thuringiensis toxin to pink bollworm (Lepidoptera:Gelechiidae) and other insects. J. Econ.Entomol. 85: 1516-1521.

tudrased by 0te u,s. Depa.hent of Agrbllur€ FOROFFICIAI- IJSE ONLY

292 vol.28 NO.4 SOUTHWESTERNENTOMOLOGIST DEC.2003 SCIENTIFICNOTE

DAMAGE, SURVIVAL, AND PARASITISMOFANTHONOMUS EUGENII (COLEOPTERA:CURCULIONIDAE) ON PIQUINPEPPER IN NORTHERNMEXICO

Luis A. Rodriguez-del-Bosqueand Marco A. Reyes-Rosas

InstitutoNacional de InvestigacionesForestales, Agricolas y Pecuarias(INIFAP), CampoExperimental Rio Bravo.Apartado Postal 172, Rio Bravo,Tam., Mdxico 88900

"Piquin" pepper(Capsicutn annuum L. var aviculare Dierb,), a wild pepper with small fruits, is beingconsidered zrs a new crop alternativefor northeastemMexico because of its increasingdemand (Rodriguez-del-Bosque et aI. 2002). Although severalinsect species,including the pepper weevll (Anthonomuseugenii Cano), have been reported feedingon piquin, damageis insignificantin most cases,especially in its naturalhabitat (Gaonaet al.2002). However,an unusualhigh infestationby A. eugenii was detectedin an experimental plot of piquin pepper during the spring of 2003 in northem Tamaulipas, Mexico, probablyas a result of populationsmoving from infestedjalapeflo plants into an untreatedadj acent plot. We presenta first report on the pepperweevil injury and survivorshipon piquin. Field and laboratory studies were conducted at the INIFAP-Campo Experimental Rio Bravo,near Rio Bravo,Tamaulipas. A 0.5-haexperimental plot of piquin pepperplanted in September2002, was surveyedfor A. eugeniidamage symptoms (Garza 2001) in 20 plants selectedrandomly on 9, 16, and 20 May 2003. Percentagedamage was estimatedby dividing the number of damagedfruit by the total numberof fruit, and multiplying by 100. During eachsampling date, 150 randomly selectedfruit showingweevil damagesymptoms were placed individually in 30-ml plastic cups to observe survival of A. eugenii tnder laboratoryconditions (27"C, 70% RH). Adult emergenceof A. eugeniior parasiteswas recordedevery 24b for 30 days.Piquin fruit not producingemerging weevils or parasites weredissected and observed using a steromicroscopeto detectother mortality factors. Dwing the periodof field samplings(9-20 May), averageminimum and maximum temperatureswer.i 24.5" and34.9'C, respectively, with a daily averageof 71.7o/oRH. First, secondand third samplingsshowed 6.9, 19.3 and 31.8%of piquin fruit damagedby l. eugenii,respectively; infestation rates never observed in the wild. Extensivesamplings of piquin pepperin its naturalhabitat in the northeasternMexican states of Coahuila,Nuevo Le6n and Tamaulipasduring 2001 and2002did not shownoticeable damage symptoms by A. eugenii(L.A.R.B, unpublished data), suggesting substantial damage to piquin may occur only in disrupted(agricultural) areas. In the laboratory,fate of A. eugeniicollected from infestedpiquin fruit in the field (n = 450)was as follows:31.6%o emerged as adults(Fig. l);4.8% reachedadulthood, but failed to emerge;33.1% died as immatures(larvae or pupae),most likely due to desiccation;24,8Yo were parasitized;and in 5.7Yo, mortality causescould not be determined. All A. eugenii adults emerged singly, in contrast to multiple feeding and emergencein larger pepperfruits (Garza200l). Parasitesemerging from A. eugenii included Catolaccushunteri (Crawford), Eurytomasp., Eupelmus sp., and Bracon sp., with 15.2,6.9, 1.8and 0.9% parasitism rates, respectively.Total parasitisrnin the presentstudy was higher than in other reportsfor 293

AUTHOR INDEX TO VOLUME 27

Aguilar,O,, 137 Jones,D.B., l3l Armstrong,J. Scott,197 Kid,P.w.,241 Amold,D.C., 152 Kindler,S.D., 131, 163 Avila-Valdez,Joel, 157 Knutson,Allen, I I Badii,M.H.,217 lkenzer, EugeneG., l2l Ball,J,, 183 Landeros-Flores,Jer6nimo, 223 B6rcenas-Ortega,N.M., 145 LaRock,D.R.,249 Barthell,John F., 97 Legaspi,J.C., 1 Bextine,Blake R., 63 Lewis,Laura R., 205 Bible,John B., l2l Liu, Tong-Xian,77, 105 Bishop,R. Jason,55 Lu,L.Y.,27 Britten,D.C.,273 Martinez,N6stor Bautista, 155 Burton,B.D.,63 Martinez-Ibarra,J.A., I 45 Camelo,L. De Azevedo,197 Mass,D.L., 152 Carpenter,J., 177 Meagher,R.L., 1 Carrillo,Tracey,249 Michels,Jr., Gerald J., 121,221, 273 Chen,T.-Y.,235 Mir, Bashir,l9 Chen,Wen, 105 Mirdad,Zohair,249 Chen,Y.P.,27 Mitchell,Forrest, 69 Chu,C.C.,235 Morrison,W.P.,241 Coates,C.,37,267 Mulrooney,Joseph E.,47, 69 Coburn,Lisa, 121 N6poles,Jesfs Romero, 155 Creamer,R., 177 Natwick,E.T.,235 dela Tone,T., 281 Nogueda-Torres,B., 145 Ellington,J.J.,249 Pair,S.D.,63 Elliott,N.c., l31, 163 Paniagua,Roberto Antonio Huerta, 155 Elzen,G.W.,255 Parajulee,M.N.,241 Elzen,P.J., 255,261 Payton,Mark E., l2l Escudero,8., 137 Peeper,Thomas F., l2l Fleenor,Scott B., 85 Piedrahita,Jorge A., l9 Fletcher,Jacqueline, 63 Pollock,Danen A.,221 Flores,A.E.,217 Olson,Jimmy K.,47,267 Foroughbackhch,R., 217 Owings,Debra A., 121,273 Gal6n-Wong,L.J.,217 Quiroz-Martinez, H,, 217 Garza-Urbina,Enrique, I 57 Raina,Ashok K., I l5 Giles,K.L., 121, l3l, 163 Ramirez-Lepe,M.,137 Godfrey,K., 183 Ramirez-Suero,M., 137 Goli,D.,69 Ramirez-Vallejo,P., 145 Gonzalez,D.,183 Rasc6n,J., 177 Guerrero-Rodriguez,Eugenio, 223 Reeves,E., 183 Hansen,James D., 205 Rendon,Pedro A., 211 Henneberry,T.J., 235, 281 Reyes-Rosas,Marco A., 293 Holmes,K.A.,69 Rodriguez-Castro,V.A., 2 17 Horn,Gerald W., l2l Rodriguez-del-Bosque,Luis A., I 57, Hranitz,John M., 97 293 Hunger,Robert M., l2l Rodriguez-L6pez,M.H., 145 Ibarra,Jorge, 2l I Royer,T.A., l2l, 131,163 Ismail,Emad A., l2l Rubink,W.,255 Jackson,Charles G., 167 Rummel,D.R.,241 Jacobi,William R., 55 S6nchez-Pefia,Sergio R., 223 Jech,L. Forlow,281 Schuster,G.L.,273 295 Shaw,R.A.,69 Simmons,Gilbert F., 205 Skow,L.C.,27 Slosser,J.E., 9l Snodgrass,Gordon L., 47 Solls-Rojas,C.,217 Southward,Morris,249 Sparks,Jr., Alton N., 77 Spencer,John P., 2l 1 Springer,T.L., 152 Taber,Stephen W., 81,85 Tao,F.L., 131 Tenorio,Felix M., 37,267 Tillman,P. Glynn,47 Yizqtez-Jaime,A.,223 Verchot,Jeanmarie, l2l Vinson,S. Bradleigh, 19,27 Wayadande,Astri, 63 Xiaodun,He, 105

296 SUBJECTINDEX TO VOLUME 27

Aceriamalherbae Coleothorpa dominicana Jrancis cana for infestationof field bindweed.273 biologyof, 9 I

Alfalfa Cotton Zygusspecies associated with, I 97 Iygus speciesassociated with, 197

Anthonomuseugenii Draeculacephalaportola damage,survival, and parasitism of distributionof in Texassugarcane, I in northernMexico. 93 Fipronil Aphis gossypii longevityof ultra-lowvolume sprays aerialmovements of anddiscovery ofon cotton,69 of malemorphs in the Texas HighPlains,241 Geocorispunctipes comparativesusceptibility of to Apis mellifera insecticides,4T comparativesusceptibility of to severalinsecticide classes. 255 Harmonia axyridis effectof spinosadon Guatemala,2ll controlof greenpeach aphids with, safetyof formic acidto for mite 249 control.261 Helicoverpazea Apples effect of temperatureand humidity arthropodspresent on removed on pheromoneproduction in, I l5 foliageof in a packingline, 205 Lygushesperus Anasatristis ovipositionby in cotton,alfalfa, and artificial feedingsystem for, 63 a wild mustard,167

Baci I lus thuringiens i s Lyguslineolaris productionof from ultraviolet comparativesusceptibility of to radiation,137 insecticides.4T

Bemisiatabaci Malathion analysesof cottonleaf characteristics longevityof ulta-lowvolume sprays effecton, 235 ofon cotton,69

Calliphoridae Megachilerotundata decompositionstudies, with a heatshock protein 70 during catologand descriptions of developmentof, 97 forensicallyimportant species of, Myzuspersicae iOeniificationof usingmitochondrial controlof with lady beetlesin the DNA,267 greenhouse,249

Circulifer tenellus Notonectairrorata incidenceof in New Mexicochile. artificial substratesfor oviposition 177 of.217

297 Orthogonisstygia Sphenophorusmaidis rangeextension, habitat, and review occuffence of easterngamagrass, of. 85 l5l

Pa ec i Iomy c e s fumo s or ous eu s Spinosad effectof substrateof on mvcosisof effect ofon honey bees in housefly larvae Guatemala,211

Pectinophoragossypiella Stethorus histrio infestationsin overwinteredand first record of from the United seededcottons. 281 States,22l

Perkins i eI I a sac c hari c i da Sugarcane distributionofin Texassugarcane, I within-field distribution of three homopteran species, I Phyllophagacrinita new pestin the Huastecasarea of Triatoma longipennis Mexico,155 genetic variability among populationsof, 145 Pinusedulis insectsassociated with black stain Trichobaris championi root diseasecenters in. 55 presentstatus of in Mexico, 153

Planococcusficus Trichograma pretiosum biologyof in California releaseof in cotton with a novel ground sprayer, I I Plutella rylostella attractionofto sexpheromone lures, Thrips tabaci r05 injury and distribution of in red cabbageheads, 77 Rhopalosiphummusae rangeextension, new hosts,new Winter Wheat morphs,and a sYnonYnrof, 8I effects of aphids, barley yellow dwarf, and grassy weeds on, 12 Sacc gar o sy dn e sac c h ar iv o r a distributionof in Texassugarcane, 1

Schizaphisgraminum economicinjury level for in Oklahomawinter wheat,163 negativebinominal distribution as a modelfor describingcounts of on wheat.131

Solenopsisinvicta artificial maturation of female alates for the production of onlY males oi 19 relatednessamong co-existing queenswith PolYgYnecolonies, 27

298 Uilted sd6 P€talSdvla€ StatemEntof Management,and 9!99!atlon October 1, 2003 Southwestern Entmologlst

4. bl$ FslHcY $20,00 Deceqber {arch, June, sePterqher, ffiV,$bb,dxt . cdnplote Mafill€ Add@ ol ffi w ALlen Knutson c/o AlIe[ Knutson Sdclety of Southweatern Entomologists, 17360 Coit Road, Dallas' Txae 75252

glais' c/o Allen Knutson SocletY of Southweatern EntoDolo 17360 Coit Road, Dallas, 1IePe 75252

Bav of Southwestern EntmologLsts' c/o Darrell Soclety Colleie Station' Texas 77843 Department of Entomlogyl-i*'" illl Untverslty' @

Darrell BaY (Mdress Above) -umairp Eonq flm anaqwle ndnQ edd@)

ro- o*m rPonaCa* bl**. ff 0E Plrffi@tur mt*ffi**w. Ziiii-tfiiwa

DepartDent of EntmologY Soclety of Southwestern Bntomloglsta Te:(as A&M Unlversity

College Statlon' TexaB 77843

fEre bx putp@t ff"il;J.Hfrfiffi ;ffi;ft-"rttb"ir|it"0"" erdtrE empi csttFtorffid g ir'Nd OwEsA Dldno Pre€dinol2lilothg (A,,ltstw otdwg€ trt0rt{ssbtqne,n) O i"" clnroea ourfng P;€dir€ i2lbdlB rcls]Mte{'/,ffi€/dotl Ps Fm 3526, oddor'1s09

301 t4. Daag Southw€atern Entmologist ftrDe 2003 Edeot lnd thtu0 of Ck{his Annge ilq Co!a.!E.dtbu 1to.coprrotttn*nh|ru- t ahg Pnedig l2 IonOE PrdbhrdXmdtoRIl|elt L a, Tohf Nutr$€rolCoplq ee$@ M) 525 525 - 346 332 b. PaH {ird R€qEtod W 0 0 Ciud.0oo - ffi*e8nffi*ffim 0 0 '4' O0tr Cl6s6 tilalt€d ltrcuC, tF t 4tps 69 68 u Enw q@|atoh tst. requq4 - ot,rzt,tsi,*i iji-'-""' ) 415 400 OrtCdFocrty 6 - DkbS!0o.t Stdedm Fm 354t 0 0 by Mal (S{rp/6, rtscffty$ Sld6d m Fm 35itl - drdmd 0 0 4,yffi otarl@) O0H Chss Mafi€d t}Iqroh the USPS 0 0 (CedbB q ofilEr mre) 4 IO Iohf Fre Dlstrtbtdbn f&m ol t'tt. qd tS..) 10 (sqo{1&,andt50 ) 419 410 Copl$ not Ot&tbqbd r06 u5 (sun ot 159.&d h.) ) 525 525

(1*, W by t'g. !n;@ tOO) 99,0 tI.) El tublo.tlon Gqufed. ASfe0on not Fql€d

October t, 2003

Instructlonsto publlshers

"* @py of u{s tom wlth }rcur postnstor annuiy e or tff$;t:.:,*[" before octobe.t . Ks€p a copy ot 1,,ecoflpleted tom ln €3€! whq€ the stoclfiolder tr .rcud.ty hddtr le a tust.e, Indud€ In lt€ms 10 afd l,l h€ whom lhe lrustee [s ac{nc. Also nm€ of 0|6 p€r8onq -rporatm fs Inctud€tie name" ano iioiiij"r oi rnai'tiiiultr.*tr *" rtckholdm or mq€ othe totalamountof bon&. mortgags, n,tlo own c hold t percent ororherriorroes ;G" i-ii.nnc *rpo.uton. h ftm box. Us€ blank shgets lf trprs 3pa€ ls rcqura. ., lf me, ch€cko,e 3. Be ilre tt fumlsh all ckcuhtlon Infottmton €llod for In ltom 15. FBe dmhton mult be shownIn ltemc lsd, c, and f. 4. It€m 15tr, Copl8 mt Dbbbut€d. mlt hdude (l) nMst4d oopl6 dStd|y Fop_35/ti, and reirned (2) edmalcd rotrms ftom nils rg€nts, and (31,eptis *gt 9n b the Frbltrhq, tq "'nrie-'rigi,i"6Gl, +or"o, um aI ohtr6pt6 mt dlr'bur€d. lf the psbll€lio had Psdodlcals luthorhadon s 6 genoEl of requestd publlcaton, hls Shtemonf ruutshcd;lt mult be pdntd h anvrJ'i;h tjJ.6ii of o|yneadllp, Mamg€rent, ifrc,lffl"X|olj1i"15 ii nnr" puuri€rrar b nor pubr$ei arme ocroo*, o. ln lt€m I 6, Indleto h€ dats of ho l3lue In wtilch thls Sht€ment of orvnershlpwilt be publlsh€d. llm 17 mustbs rlgn€d. Fdlun',oflb orpttbrithttr,bt *,toro.,oro.sttpnay bdtoarwndonor p.tt,'it"drrtthqto,tbn. PS Fm t526, Octobet1€re/g (Rewset

302