~q February 1994 HA

The purpose of theitor and evaluate tholiothil virescem (F.),Helicoverpa ua (~b'eatment with an Hrmine if and to what e)tions of this pest comb'eatmenL Here we (mone trap counts froling) generation througtion F i; we comparespecies in the b'eate.with the b'eated plotmenL Results for Ficotton and other hostways to improve effe(

For H. virelcens antmost valuable measucurrent economic threon egg numbers (calcuminals infested; e.g., ~Mississippi CooperatHowever, obtaining atlaborious and impractiiog the early part ofanalysis is complicate<host attractiveness. Fhigh because plantingand developmental phsoil and weather corwidely at this time (reaches the pinhead-~ment. oviposition by ~

relatively rare (Hayes cman 1989, Hayes 1990)solinaceous species anseason cotton, Abutil.(velvetleaf or wild corovipositing females. Imore attractive and C1nates much of the vel,creasingly for ovipositi

Although we cannot I:to capture numbers on?D1one traps are reliablltaring H. virelcens alproblem with use of piapProPriate life table d.begg or larval densities

rs (Fitt 1989). Howe

.r;:sperforrnance (Hayc199 ; Hayes II; Colel

1 b; Witt et aJ. 1~

~p capture data accula undance and fluctw.

~e POpulations. Whe'laVe. Lte lnnerent biases, tUS~ than sampling effo

of pheromone trap c(

ti suppression achieve<~n ~f HNPV. Egg COUI

ertze the surviving

BIOLOGICAL AND MICROBIAL CONTROL

Evaluation of Early-Season Baculovirus Treatment forSuppression of Heliothis virescens and Helicoverpa

zea (Lepidoptera: Noctuidae) Over a Wide Area

JANE LESLIE HAYES. AND MARION BELL-

J. £eon. Entomol. 87(1): 58-66 (1994)ABSTRACT Pheromone trap counts of F I male cotton bollwonn. HelkO1Jef'JHJ UG (Bod-die), and tobacco budwonn. Heliothi8 oirelceRl (F.), were used to usess the eft'ect ofareawide suppression Kbieved by early-season application of a HelkOWrpGiHeliothi8-speci6c nuclear polyhedrosis virus. Eas (F,) were collected &om cotton and od1er hoststo characterize the surviving reproductive populations. Trap and egg collection sites wereestablished at 1.6-kIn Intervals (n - 5) in four cardinal directions &om the center of controland treated plots (259 kin', -10 by 10 mi). Traps also were placed at 6ve additionalintervals in four cardinal directions beyond the treated p~ to usess the e«ect of dispersal.The eft"ect of treatment was demonstrated by deviations In trap capture patterns within a ,year between treated and control plots and between years In the treated plot. Rates ofincrease between generations were calculated &om the number of moths captured in onegeneration divided by the number &om the previous generation. The rate of increase forthe first 6eld generation of H. oirelcenaln the treated plot (1M) wu 13~ comP8red with38~ in the control plot (1~) and 38~ In the treated plot In the year before treatment(1989). The rate of increase for the 6rst 6eld generation of H. UG wu ~ in the treatedplot (1M) compared with 55~ in the control plot (1M) and 95~ in the treated plot theyear before treatment (1989). Rates of increase for both species in the subsequent gener-ation remained low In the treated plot (1M) compared with the control plot (1M) andwith the previous year (1989). Our results indicate that a single virus application canreduce the adult H. oirescena and H. UG populations emerging &om alternative hostspresent early in the season. Other methods for improving the eftlcacy of treatments are

discussed.

KEY WORDS Heliot1ril/Hel~, baculovirus, area-wide evaluation

EVALUA110N OF THE SUCCESS of suppressiontactics over large areas presents serious chal-lenges to standard small-plot or 6eld-size exper-imental procedures. The task is puticularlydifficult in an open and heterogeneous ag:roeco-system, where the targets are highly polypha-gous and mobile like the memben of theHeliothi8IHelicoverpG~mplex. Available fund-ing and practical considerations (operational andlogistical) preclude the level of replication ne-cessitated by the presumed plot-to-plot variabil-ity. For such large-scale entomological studies,Schneider (1989) reviewed three approaches.The optimal approach was a ~yr study with re-venal of treabnent and control sites in alternateyean. In this approach, both year and locationserved as replicates or, more accurately, stan-

dams for comparison. Other investigations, suchas watershed studies, commonly use historicaldata and reference sites as standards for compar"

ison.The evaluation design for the Heliothis nu-

clear polyhedrosis virus (HNPV) pilot test dis-cussed by Bell &: Hayes (1994) followed the de-sign described by Schneider (1989). In 1900, tWO259-km2 (16 by 16 km) plots were established inan intensive cotton production area in the deltaregion of Mississippi. One plot was designatedthe control and the other plot was designated asthe treatment plot for year 1. In principle, thesedesignations would be reversed in year 2 of theexperiment. However, historical data were avail-able for the area designated as the treatment plotfrom a long-term movement study conducted ill

approximately the same location from 1987-1yS9(Hayes 1900). In the Bell &: Hayes study (1994),both male adult counts from routine pherornonftrap S;aJnpling and egg counts from inspecti?nslOsterminals of cotton were taken. These vanab twere used to assess the effect of the HNPV tre,t-

ment in 1900 (Hayes 1990).

- -This article reJM)rts-die results Or rese8n:h only. Mention of.

proprietary product does not constitute an endonement or .

recommendation rOl' its use by USDA.I Southern Forest Experiment St8tion. FOIest Service,

USDA. Pineville, LA 71..I Southern Insect Manapment I..boratory, USDA-ARS,

Stoneville, MS 38776.

February 1994 HAYES &:: BEl AREAWIDE EVALUAnON OF Heliothi.r NPV TREAnoIENT 59

~atment for~licot;erpade Area

(e.g., host, insect species composition, and dis-tribution). The effect of treatment can be demon-strated by deviations in trap capture patterns be-tween treatment and control plots within a yearand between years in the treabnent plot.

Long-range and mesoscale movement of H.virescen.r and H. zeD are significant in areawidecontrol (Knipling & Stadelbacher 1983, Schnei-der et aI. 1989). For the treabnent plot, we ex-tended sampling of both adults and eggs beyondthe treabnent boundary to assess the effect ofdispersal. Previous studies by Schneider et al.(1989) and Hayes (1991a) showed that H.virescen.r and H. zeD can move as far as 20 kmand can typically move 3-8 km per generation,depending on environmental conditions. Thus,we expected that HNPV application would bemost apparent in the center of the treatment plotand would dissipate at the borders. However,spatial heterogeneity in the area is high andcould obliterate the predicted effect of treatment.For both reasons, detectable suppression waslikely to be Iwomped out from one generation tothe next.

'84

sess the effect ofol;t'rpuiHeliothis-n itnd other hostsl~ction sites were~ center of controlat five additional~lfect of dispersal.patterns within Iited plot. Rates ofIS captured in oneate of increase for1% compared withbefore treatment

16% in the treatedIe treated plot the,ubsequent gener-01 plot (1990) andIS application canI alternative hostsof treatments are

Materials and Methods

Preparation, application, direct evaluation ofHNPV perfonnance, a general description of thestudy area. and speci6c attributes of the treat-ment and conb'ol plots are described elsewhere(Bell &: Hayes 1004). Only those procedures spe-cific to the areawide field evaluation of the studyare described in detail here.

Study Desip. In both treatment and conb'olplots, at least four sampling sites were estab-lished within each 1.6-km interval (radius) fromthe center (Fig. 1). On the basis of availability ofaccessible cotton fields, we spaced samplingsites per interval in different quadrants corre-sponding to cardinal points. In the control plot,sampling extended to 8 km (6ve intervals). In thetreatment plot, sampling was extended to 16 kin(10 intervals) to pennit us to assess the effect ofmigration on the treated area. (Hereafter, inter-vals 1-5 are referred to as the treated subplot and6-10 as the untreated subplot.) At each site, twotraps (one trap per species) separated by -100 mwere placed at the edge of a cotton field alongan accessible roadway. Because fields are fre-quently cultivated and roadsides are oftenmowed or burned. traps were placed near powerpoles and in other protected sites. The nearbycotton 6eld(s) was routinely searched for eggs.

Moth Sampling. Standard 75-.50 hardwarecloth, cone-shaped pheromone traps (Hartstacket al. 1979) were used to monitor the relativeabundance and ftuctuations in the adult male H.vire.rcem and H. zed populations in the treat-ment and conb'ol plots. Traps were routinelymonitored from 1 April to 1 August (day of year91 to 212). This period encompassed the ftights

Other investigations, suchcommonly use historical

~s as standards for com par-

;gn for the Heliothis nu-IS (HNPV) pilot test dis-s (1994) followed the de-eider (1989). In 1990, twoplots were established inoduction area in the deltaOne plot was designated,er plot was designated asyear 1. In principle, thesee reversed in year 2 of the. historical data were avail-nated as the treatment plotement study conducted inIe location from 1987-1989Bell &: Hayes study (1994),ts from routine pheromonecounts from inspections of

ere taken. These variablesIe effect of the HNPV treat-1900).

The purpose of the cum-nt study was to mon-itor and evaluate the tobacco budwonn. He-/iothis virescens (F.). and the cotton bollwonn.Helicot,'erpa :.ea (Boddie), populations aftertreatment with an HNPV. We sought to deter-mine if and to what extent the areawide popula-tions of this pest complex were affected by thetreatment. Here we describe results of phero-ntone trap counts from the parent (overwinter-ing) generation through the second field genera-tion F 2; we compare rates of increase of both!opecies in the treated plot with a control andwith the treated plot in the year before treat-ment. Results for F 2 egg collections made oncotton and other hosts are reported. We discussways to improve effectiveness of the treatment.

For H. vire8cens and H. zed, egg counts are themost valuable measure of abundance becausecurrent economic thresholds on cotton are basedon egg numbers (calculated as percentage ofter-minals infested; e.g., Mississippi Conb'ol Guide,Mississippi Cooperative Extension Service).However, obtaining adequate infestation data islaborious and impractical on a large scale. Dur-ing the early part of the growing season, dataanalysis is complicated by the rapid changes inhost attractiveness. Field-to-6eld variability ishigh because planting times vary among growersand developmental phenology is dependent onsoil and weather conditions, which 8uctuatewidely at this time of the year. Until cottonreaches the pinhead-square stage of develop-ment. oviposition by H. vire8Cen& or H. zed isrelatively rare (Hayes et al. 1988, Hayes &:: Cole-man 1989, Hayes 1900). Simultaneously, anothersolinaceous species and common weed in early-season cotton, Abutilon theophratji Medikus(velvedeaf or wild cotton), is used heavily byovipositing females. Later, as cotton becomesmore attractive and cultivation of fields elimi-nates much of the velvedeaf, cotton is used in-creasingly for oviposition.

Although we cannot ascribe an economic valueto capture numbers on a day-to-day basis, pher-omone b'aps are reliable sampling tools for mon-itoring H. vjre8c6m and H. zea. The primaryproblem with use of pheromone b'aps is lack ofappropriate life table data to pennit estimation ofegg or larval densities from male capture num-bers (Fitt 1989). However, extensive studies oftrap perfonnance (Hayes et aI. 1988; Lopez et aI.1988; Hayes &:: Coleman 1989; Hayes 1900,1991b; Witz et al. 1990, 1991) have shown thattrap capture data accurately re8ect the relativeabundance and 8uctuations in the local adultmale populations. Whereas all trapping deviceshave inherent biases, they are far more consis-tent than sampling efforts by humans. Thus, weused pheromone nap counts to assess the degreeof suppression achieved by early-season applica-tion ofHNPV. Egg count data were used to char-acterize the surviving reproductive population

February 199460 Vol. 87, no. 1JOURNAL OF ECONOMIC ENTOMOLOGY

eggs) or a proportion (not <20'*' of eggs for sam-ples >20) were placed on an arti6cial diet forrearing to the adult stage for species identi6ca-tion. If21-100 eggs were collected, 20 eggs wereplaced on the diet. If 101-125 eggs were col-lected, 25 eggs were placed on the diet. If 126-ISO were collected, 30 eggs were placed on thediet, and so on. Species composition of thesesubsamples were used to estimate the proportionof H. oirelcen.r and H. zeD per sample.

D8ta Analysis. Rate of increase (rJ betweengenerations was calculated by taking the naturallog of the total number of moths captured in ageneration (nJ and dividing the resulting num-ber by the total number of moths captured in theprevious generation (n(l - I). Generations (P -no. F 1= nl, F. - na) were de6ned with both eggcollection and trap collection data, i.e., by in-cluding the duration of the oviposition periodand the corresponding period of trap count in-crease.

~ CONTROL~ \1 PLOT

H. viresc125~ .,. Treo'

I HWY 81 .W-t--I

.-m~ 100:>-0

0

~ 75c~0uc 500.2

25.

r,=O.13rfO.70:§~~~!~ 82

~~:";\~~

~"D~TREATEDI 'I.,_~

ITONEVlllE. MS.

Fia. 1. Schematic di8lr8D1 of l~ HNPV early-season treatment study area - Stoneville, MS. aadevaluation design; abundaDce of H.IWtlau/H.lico-t)e'fpCI adults and ellS wu monitored at four locationsper interval (l.6-km radius) &om d1e center of die COD-trol and treatment (treated and untreated subplots)plots.

OJ. .

125rs:- Unir

100

75

50

25

',=0.51,.,0.66

)..0

0"".-c~0u

ca.:2

OJ. .125-

100

75.

50~

25

).0

Q"...c~0

(Jc0.~

r ,=0.3r f 1.(

of the emergent overwintering or migrant paren-tal generation (P) and Hrst two Held generations(F I and Fa>. Using the method described byHayes (1991a), we emptied traps three times perweek (Monday, Wednesday, Friday), at whichtime captured moths were counted and recorded.To estimate mean per trap count per day, trapcollections were divided by the number of daysof operation between collections. Pheromonelures (Zealure, Hercon Environmental; Virelure,Scentry, Buckeye, AZ) were replaced biweekly.Traps were repaired as needed and the area un-der each trap was kept clear of vegetation.

Ea SampliDa. In both control and treatmentplots, one to three cotton Helds near each traplocation were searched routinely for F. eggs (andlarvae, if present) during the Sight of the F Iadults. We sampled multiple sites per location toameliorate the site-to-site variability in crop phe-nology. The developmental stages of cottonranged from cotyledon to pinhead square whensampling began and from pinhead square toSowering when sampling ended. Velvetleaf wasinitially prevalent in many fields and was sam-pled along with cotton. Some sites containedcom or velvetleaf only and were sampled to as.sess the overall rate of oviposition and speciescomposition in each plot more accurately.

Each field was searched by one or more fieldcrew memben (who inspected terminals at ran-dom) for 30 min at intervals of 2-3 d (data wereadjusted to account for number of samplers pervisit and frequency of visits per site where ap-propriate). Crop growth stage was recorded ateach visit. All eggs (or larvae) encountered weret'ollected in 3O-ml plastic cups, kept separate bylocation and host. and returned to the laboratoryfor processing. Eggs were counted and all (if <20

OJ.--:-~-75 95

Results md Diseussion

Moth Samplinlo Mean trap counts per day forboth species over the 1~ sampling period (P,F I' and F 2 generations) from the control andb'eabnent (b'eated and untreated subplots) plotsare shown in Fig. 2. Typically, we observed phe-nolOKical diHerences between the two moth spe-cies across plots (e.g., 1~) (Fig. 3). For the F 1 orb'eated generation (day of year 137 to 177), asharp increase in trap captures of H. zea beganon DOY 152 in all plots, whereas trap captures ofH. vire.cen.t showed a less dramatic increase10 d later (day of year 162) (Fig. 2).

We observed diHerences in species composi-tion. In the control plot, H. zea was the predom-inant species (>75% of all moths captured duringthe F 1 generation). In the b'eatment plot, thecount was nearly equal (54% in treated subplot,49% in untreated subplot). Exemplifying one ofthe problems with conducting controlled fieldmals at this scale, these diWerences in speciescomposition between plots were expected giventhe land area involved. The control plot is morenortherly and is situated in closer proximity tothe Mississippi River than the b'eatment plot.Although the distance between the centers of thetwo plots was only -30 km, the combiRation ofweather and soil differences may have been suf-ficient to account for the observed differences inboth host phenologies and H. vire.cen.r:H. zeDratios.

Regardless of proportional diHerences in spe-cies composition, trap captures for both speciesincrease with each successive generation untilmidseason under nom1al conditions (e.g., 1989,Fig. 3). From the parental to F 1 generations,mean captures per day may increase twofold.Visual inspection of the trap capture patterns forthe sampling period revealed deviations from ex-

Filo 2. Meanpheromone trapsof H. UG males(e) untreated subM In n/lnN. - 1 '

62 HJOURNAL OF ECONOMIC ENTOMOLOGY Vol. 87, no. 1 February 1994

T~. T~C8i-C8/s.

~

was 13% compared with 38% for the conb'ol plotand 51% for the untreated subplot. The rate ofincrease for H. zea in the treated 3ubplot was36% compared with 55% for the conb'ol plot and67% for the unb"eated subplot.

For both species in 1989 (Fig. 3), the rate ofincrease between the second and third genera-tions (,.to> remained high; H. vire,cem showed asubstantial increase which coincided with thetypically observed high field infestation around4 July. The 1900 results show comparably high"2values for both species in the control plot and inthe untreated subplot for H. zea (Fig. 2). In thetreated subplot area, the change in ,. values("1 versus ".) is comparable (i.e., -60% for H.vire,cem and 15-20% in H. zea); however, nei-ther species appeared to recover completelyfrom the loss during the next generation (i.e., H.vire,cem in the conb'ol plot showed a 104% rateof increase versus 70% in the treated subplot; H.zea in the conb'ol plot had a 95% increase versus48% in the treated subplot). Relatively low val-ues in the untreated subplot (especially for H.vire,cem) may reSect the inSuence of the neigh-

- 800,c~0

u .00.

g:... ~o.-0"0t- 200

H. viresc..,.. "

2501 ...! .- 1 ! ~ 2 j '.122' .

>0.2000

0175

~150.C . I~ 1251 "1 =0.38 i

IUI°O 1 "2=1.63 i,C :!

0 75 :.2 50

25

_I.""'-=::;,SO '40 ,, 000~I;~.-- l~

-800c~Co) _.g~ .00.0'00- 200.

~\oJ- I_~.~~~~~~~~ .

80 100 120 140 110 180 200 220 2.0 210 280

250 H. %.(1. .2251 P! f, ! f21 I

>-2000

0175

C'50S'25U'00c0 75.2 50

25

r, =0.95r2=0.71

I. vir.scans . .moths o. eggs0

y a -3.54X + 45.8 A. " A' - <0.01 0 0'0 ' "

'"

50~'.0130

!~~J~;J80 100 120 140 160 180 200 220 240 210 280

Day of Vear

FiCo 3. Mean number per day of (a) H. ~and (b) H. %eO males captured in 1989 (DOY 91-212) inpheromone traps in d1e area tIeated by HNPV in 1990.Rate of increase (rJ for the 6rst and second generations(calculated as In n/lnn, - 1 where n, is number of modisper generation, P, F I' and Fa> are given.

40.FiCo 5. Total numb

site) of eggs (Fz) coUethe Hight period of F 1(B) untteated subplot, .

~~.

", Y - 0.096)( + 23.U" I. . ',... R' - 0.34 I

",po

boring b'eated subpl(into the unb'eated su

The possible effecparent in the phero(Fig. 4). Neither lineanalysis indicated stween mean capturetance from the centeresult is not surprisispatial heterogeneit:degree of temporalthe elongated applic

Ea: Sampling. Hftectable level of supHNPV b'eatment. Wf;(F 2 generation) to prlabout the reprodu((treated) generationand abundance of 0'tion for the-control aparent in the compapie site over the sarwere detected firstearlier in the b'eatlthan in the control pIin local planting cc

tI)~.- 20.

"CG

~ 10.--> .G 0

C 0; 2 ; 4 i i 7 8 9 10--E 5O.H. z-

5 BU401c:301~

y - .o.87X + 30.5. RJ - 0.22

20-

100

pected catch in the neated subplot (neated in1990) compared with the previous year (1989)and with the control plot and unneated subplotfor 1990 (Figs. 2 and 3). In the neated subplot, H.zeo initially increased during the F 1 generation;however, F 1 H. vire8cens failed to reach a meancapture rate per day that exceeded the parentalgeneration peak. At the same time, in the adja-cent untreated subplot, both H. vire8cens and H.zeo patterns appeared normal (i.e., peak meancaptures per day during the F 1 generation ex-ceeded the parental generation peaks). In thecontrol plot, the mean captures per day peakexceeded the parental generation peak and con-tinued to climb after the treated and unneatedsubplot values showed a decline.

Calculations of rate of increase (rJ confirmedour visual assessments of trap capture data (Figs.2 and 3). In general, the rate of increase for bothspecies in all plots in 1990 was lower than thosemeasured for 1989; the rate of increase for H. zeoin both years was higher than that for H.vire8cens. Both species had lower rates of in-crease in the treated subplot compared with theuntreated subplot and control plot. The rate ofincrease for H. vire$cens in the treated subplot

.Y - -0. 137X + 2.8

. AI -O.~0~ - -~- -;- - -i --; - - ~- - f - -, - -, - -;0

Radius From Center (1 .6 km)Fia. 4. Frequency of modts (mean number trapped

per day; solid circles) and eggs (mean number per siteper day; open circles) for: (A) H. vire8CenI and (b) H.zeo collected at intervals (l.6-km radius) &om center oftreabnent plot. Line 6tted by regression (dashed-line,moths; solid line, eggs): for H. mre8Cenl moths, r -0.04209 and eggs, ,. = 0.5852, and for H. zeo moths,r - 0.4684 and eggs, ,.a = 0.3(XX).

laG 1.0 I

10001c.~ -800 ci<->800

63HAYES &: BELL: AREAWIDE EVALUA110N OF Helwthil NPV TREATMENTFebroaIY 1994Vol. 87. no. 1

'000 ISOA. T ~ .. w.../H. --a 800 CN I 2& ,.,

~ . T~c.. lOO~.3 800 0 c.-,. n

g 7S§"",.00 -"0 SO '0 ~

.- 200 2S-

0 0130 ~ 1.0 leo

'000 ! ISO

8.~ /.- 800 T\ '2S,.,§ 1\ oa0 J ; lOOoa

(,.)800 \ n

g I. ; 7S §,.00 II' \ -

"0 'I SO"""'0 I ~.- 200. J [ 2:-

rlte control plot_lot. The rate of:00 subplot wascontrol plot and

A. T r.-d Vlhtt*f

. H . H.-

~

.120ji7, 1 00.- 80,C~

~&O, ~~ 4Q,~

j~ ~..#~~~ ~l 20

0 - ..#~ -

1S0 1.0 180 1eo 170 1eo

1.o.~ I'a.u... I

120,.ain 1 00 '

_eoc~0 80,V~~ .0'

,- ~ ~:..-"~-,j ~ ~ 20 .. ,... ,

0 . 11SO 1.0 1SO 1eo - 170 t.

t-lt. 3), the rate ofld third genera-1scens showed aIcided with thefestation aroundnparably high r2nb-ol plot and in~ (Fig. 2). In thege in r values~.. ~60% for H.); however, nei-Iver completelyrleration (i.e., H.Ned a 104% rateated subplot.; H.increase versus

latively low val-specially for H.Ice of the neigh-

)L- ~t1,h~' : . r'_I . .. 6 1'/\ r..,.

1SO 1.0 150 180 170 180

1000, .150

c.-. 125,.,

100:8n

752a.

so-l/)==

258

0

tC88

A1

- 800c~(,,) 800g !... .-1"0-00- zoo l-J\.-

Iths o. eggsl

0 AI

~I.I\~I. -~ ~ l' -~ -:--:

-130 1.0 150 180 170 180

Day of V..

Fi.. 5. Total number and frequency (mean count/site) of eggs (F,) collected &om cotton fields duringthe Right period of F 1 adults in the (A) control plot,(8) untreated subplot, and (C) treated subplot.

= 0.098)( +.23.J

'_O.~ -_:J

120.

,:11'1100

- 80.c~0 80.

Co)

~.o... '- -~- () ~~ ~ 20, ) i , .' rII -- -130 140 150 180 170 180

_afT.Fi.. 6. Frequency of oviposition by H. oirelCeM

and H. ~ on velvedeaf in: (A) control plot. (8) un-treated subplot. and (C) treated subplot.

8 9 10

B

.87X + 30.5

.22

~- - -~- - -- - -!7 8 9 10r(1.6km)n number trappedn number per site"elcen6 and (b) H.ius) &om center ofsion (dashed-line.~Cenl moths, ,2 =for H. %eO moths,

Vol. 87, no. 1 Febroary 199-64 JOURNAL OF ECONOMIC ENTOMOLOGY

,,

...c,~0uc~G~

150 ,,~. ::~~:d H. Vir..c., ,~. Treated H. vir..c.

125"E . moth.~ 100. . egg.(..) 75.C0~ 50.

:225.

A.T~

.H.~°H.-

CoIIaII A. Tr.-1~in

C10,~

U ;

GO ,GO a.

...

. "'

. e,

.--~~~h-,. 1\'\~~ 0'-70 1

150,B. U

125...C100.j0U 75.

Cca 50~~ 25

13020

'.0 lag 1ao 170 ,eo '.0 zzo100 130

.-,~.in

C'0.~0

<..)atQ

...

12S,-C~ 1000

U 75C0. SO.~

2S. ~i .I\- I0" ecl=--=~-__l~' '"- ~~ .I70 '00 '30 '.0 '.0 220

130 1.0 70,~O leo 170 leo

c.~ c c. Cont,.ol.-,§.in'E 10.

~0Co)COCO

,"E~10

UC0~

~,..,

.-" ,l~ .

125,..

§100~0 :

U 75;

Ce 50.G~ 25;

0'-70 .'30 '.0 ,so '.0 '70 '80

~ofY..Fig. 7. Frequency of oviposition by H. vire.tceft.r

and H. zea on cultivated cotton in: (A) control plot.(B) unb"eated subplot, and (C) treated subplot.

Fis. 9. Me!hired and e~in the: (A) con"treated subplot

I ~RL It!I" I~a_~~ ~ .I

70 '00 '30 '.0 '.0 220Day of YB

FiCo 8. Mean number per day of H. virelceM malescaptured and eggs collected during the sampling pe-riod in the: (A) control plot, (8) unheated subplot, and(C) treated subplot.

In summarsingle virusvirescens andalternative h(though the npopulation asof preliminarare encouragthe overall Slspray covera,coverage w~ditions durinhigh winds, ations were aIwith the restended for 2]liest in the sfore neatmewould therelwhy the H. %appeared to tof increase I::emerge in tlbaculovirus .laid are hatci

away from cultivated cotton and into a relativelyeasily managed situation (e.g., with limited pes-ticide or herbicide treatment and mowing).

Comparison of egg and trap count data (Figs. 8and 9) confinued previously observed temporalrelationships (Hayes et aI. 1988, Hayes &; Cole-man 1989, Hayes 1990, Witz et aI. 1990). Thesignificant correlation coefficients for H. zea eggand trap counts over time were obtained by com-parison of same day (egg DOY = trap DOY) orI-d lag (egg DaY - trap DOY 1), suggesting thatpheromone traps might be useful devices to in-dicate an increase in number of damaging imma-tures of this species within a field. However,significant correlations for H. virescens were ob-tained with a delay of7 and 8 d (egg DaY = trapDOY 7 or 8); the highest coefficient was obtainedat a delay of 8 d in the treated subplot (r =0.8235; n = 9; P < 0.0064). Thus, trap collectionsshowed a corresponding increase in moth num-bers a full week or more after egg counts beganto increase; therefore, traps do not providetimely infonuation for prediction of infestationsofH. virescens.

The obvious predominance of H. vire&cens re-vealed by rearing is consistent with results fromprevious studies in this area (e.g., Hayes 1991b).Despite the capture of relatively high numbersof H. zea in pheromone naps (i.e., often equal toor higher than H. vire8cens), the incidence ofoviposition by H. zea remained significantlylower than H. vire&cens in cotton fields through-out the growing season. Previous studies showthat this difference in oviposition persists in sub-sequent generations and throughout the growingseason (Hayes 1990, 1991b). Differences in com-position of adult (male) between plots was not asapparent in species composition of eggs col-lected from nearby fields (14% versus 10% H. zeain control versus treatment plots).

The high incidence of oviposition on vel-vetleaf by H. vire8Cen& (and to a lesser extentH. zea [Fig. 6]) suggests that velvetleaf hashigh potential for population management. Ap-parently, well-timed cultivation would contrib-ute substantially to suppression of moth numbersin subsequent generations. More sophisticatedtrap crop systems may be designed to take ad-vantage of this attractive host (at least during thistime frame) and to draw substantial oviposition

l

50

!25 00

75

SO-

25.

IZ5i

looj75

50

Z5.

Vol. 87, no. Mf"ebnJalY 1994 HAYES &:: BELL: AREAWIDE EVALUA110N OF Heltothu NPV TREAnI.ENT

~;;;;;;;~laid are in the last instal. Use of additional air-craft may dramatically decrease the applicationtime. Unfortunately, poor application condition.ccannot be overcome as simply, but could be off-set by multiple rather than single application;doubling the application rate would ensure im-provement in coverage.

These results also have important implicationsfor evaluation technology, particularly the valueof pheromone traps for this purpose. Trap cali-bration still has serious problems, but trap cap-ture data appear to reSect the 8uCtuations in thepopulation abundance necessary to monitor re-sponse to b'eatments. Additional studies of adultdemography (dispersal and survivorship) areneeded to unravel the relationship between ovi-position and male 8ight. Finally, the high inci-dence of oviposition on velvetleaf versus cottonsuggests a second and environmentally soundsuppression tactic to compliment the use ofHNPV.

,./"'1"\~ '.0 ~

220

~~~~~Jleo aao-,

,so A. T,.eated H. .em,,~ . moths

~,oo . egg.

(.) 7~

;so~~ ,~

070 '00 ,so ,eo

'SO'-- I'B. Unt,.eated'2~

:,00

U 7-

! 80

' ~ ~ I ,.070 '00 ,SO 1eo 180 .-'801 c. Control - .

'28-~ '00-I

7-

AcknowledllDenbWe thank the memben of the labontory and field

crews of the Southern Insect Management Labontory,USDA-ARS, Stoneville, MS, for their hard work underoften adverse conditions. Special thanks to Don Hub-bard (Southern Insect Management Labontory,USDA-ARS) for his extnmdinary e~ in coordinatinlfield sampling and to Faye Guinn (Southern ForestExperiment Station, Forest Service, USDA) for prepa-ration ofb pphics. We appreciate the helpfuI com-ments on earlier dnfts of dte manuscript provided byJ. L. Robertson (Plcl&c Southwestern Experiment Sta-tion, USDA-FS) and P. B. Turchin and B. L. Strom(Southern Forest Experiment Station, Forest Service,USDA).

~~~'.0 I

220

, : /__~~,.l~~ r ~ I

70 too t~ t80 teo 220

~ofY-Filo 9. Mean number per day of H. "G males cap-

tured and egs collected during the sampling periodin the: (A) conb'Ol plot, (B) untteated subplot, and (C)treated subplot.

'irelcmI males! sampling pe..~d ~ubplot, and

In summary, our results indicated that thesingle virus application reduced the adult H.vire8cens and H. zea populations emerging fromalternative hosts present early in the season. Al-though the b'eabnent failed to reduce the adultpopulation as much as was expected on the basisof preliminary small-scale studies, these resultsare encouraging. The primary factors aJfectingthe overall success of the treabnent were lack ofspray coverage and timing of application. Spraycoverage was adversely aJfected by windy con-ditions during the application period. Because ofhigh winds, appropriate aerial application condi-tions were also limited to narrow time periods,with the result that the application period ex-tended for 21 d. During diis time, eggs laid ear-liest in the season would reach pupal stage be-fore treabnent reached all areas of plot andwould thereby escape. This pattern may explainwhy die H. ua population in die b'eated subplotappeared to experience a sudden drop in the rateof increase because this species was the 6rst toemerge in the plot. Ideally, the application ofbaculovirus would be made when the last eggslaid are hatching and the larvae of the first eggs

0 a relativelyI limited pes-1owing).: data (Figs. 8,'ed temporal,yes ~ Cole-

l~). Ther H. zeo eggned by com-ap DOY) or:gesting that~vices to in-

laging imma-d. However,em were ob-: DOY = trapwas obtained;ubplot (r -p collections1 moth num-:Ounts begannot provider infestations

References Cited

Bell, M. R. &: D. D. Hardee. 1991. Spraye8iciencyof aerial application of nuclear polyhedrosil virus inarea-wide treabDeDt of early season hosts of boll-WO~udwOl1DS. pp. 624-~. 'n Proceedings,Beltwide Cotton PnxIuc:tion ~ Conference,7-12 Januuy 1991. San Antonio, TX.

Bell, M. R., D. D. Hardee, J. L Hayes &: E. A. Stadel-b8Cher. 1991. Manacement of cotton bollwonnand tobacco budwonn populations duough area-wide application of nuc. polyhedJOlil virus onearly-seuoo alternate hosts, pp. 24-37 .'n R. Soper,N. Filippov &: S. Alimukbamedov [edI.], Cotton-integrated pest management: Proceedings, Sympo-sium, 3-Q September 1~. Tashkent, Uzbekistan,USSR. USDA-ARS Publication ARS-1~. USDA,Washinlton. D.C.

Bell, M. R. &: J. L Hayes. 184. Area-wide manage-ment of cotton bollworm and tobacco budwonnthrough application of a nuclear polyhedrasil viruson early-season altemate hosts. J. Econ. Entomol.87:~.

Fitt, C. P. 1_. The ecolocy of d1e H.liothil spe-cies in relation to agroeoosystems. Annu. Rev. En-tomol. 34; 17-S2.

Vol. 87, no. 166 JOURNAL OF ECONOMIC ENTOMOLOGY

Leaf Con~(Co

ABSTRACthe effectVuillemin,decemline.causing tiTto any ofphagostimrates on dainoculatedfungus rabconsumptithe three I

I(:EY WOf

1988. New Orleans, LA. National Cotton Council ofAmerica, Memphis, TN.

KnipUnc. E. F. &: E. A. Stadelb8eher. 1M3. Therationale for area-wide management of Heliothu(Lepidoptera: Noctuidae) populations. Bull. Ento-mol. Soc. Am. 29: 29-37.

Luter, M. 1.., W. F. Kitten, E. F. Kniplins. D. F. Mar-tin, J. C. Schneider &: J. W. Smith. 1987. Esti-mates of overwintered population density and adultsurvival rates of Heliothu oirelceM (Lepidoptera:Noctuidae) in the Mississippi Delta. Environ. En-tomol. 16: 1076-1081.

Lopez, J. D.. J. I.. Goodeooulil. T. N. Shaver. J. A.Witz, B. A. Leoabardt &: W. A. Dickenon. 1988.Potential use of sex pheromone traps for monitoringHeliothil zeD and oirelcem, paper no. G4. In Pr0-ceedings, Beltwide Cotton Production ResearchConference, 3-6 Jan. 1988. New Orleans, LA. Na-tional Cotton Council of America, Memphis, TN.

SchDeider. J. C. 1_. Role of movement in evalua-tion of area-wide insect pest management tactics.Environ. Entomol. 18: 868-874.

Schneider. J. C., R. T. Roush, W. T. Kittea 6: M. I..Luter. 1989. Movement of HeliothU oirelcem(Lepidoptera: Noctuidae) in Mississippi in thespring: implications for area-wide managementEnviron. Entomol. 18: 438-~.

SAS Institute. 1985. SAS users' guide: statistics,version 5. ed. SAS Institute, Cary, N.C.

Witz, J. A.. J. I.. Hayes. M. A. L8tbee£. J. D. Lopez &:F. 1. Proshold. 1990. Use of pheromone trapsand the TEXCIM model to predict HeliothU spp.egg populations in cotton. Southwest Entomol. 15:217-229.

Hlltstack, A. W., J. A. Witz &: D. P. Buck. 1979.Moth b'aps for die tobacco budwonn. J. Econ. En-tomol. 72: 519-522. ,

Hutst8ck, A. W., J. A. WiIz, J. P. HoUinpworth, D. P.Buck, J. D. Lopez &: D. E. Hendricks. 1978. Re-lation of tobacco budwonn catches in pheromonenaps to field populations. Soudiwest. Entomol. 3:43-51.

Hayes, J. L 1988. A comparative study of adultemergence phenologies of Heliotlall oitW~ (F.)and H. rAG (Boddie) (Lepidoptera: Noctuidae) onvarious hosts in field cages. Environ. Entomol. 17:344-349.

Hayes, J. L 18). Relating HeliotJau spp. phero-mone b'ap captures to egg counts in cotton: d1irdyear data &om dle delta of Mississippi, pp. 198-D.In Proceedings. Beltwide Cotton Production Re-search Conference. 9-14 January 1990. Las Vegas.NY. National Cotton Council of America, Memphis.TN.

Hayes,J. L 1991.. Elemental marking in arthropodpests in agricultural systems: single and multigen-erational marking. Soudlwest. Entomol. 14: 37-47.

Hayes,J. L. 1991b. Dynamics ofnoctumal behaviorof adult HeliotJau spp. (Lepidoptera: Noctuidae) incotton. J. Econ. Entomol. 84: 855-865.

Hayes, J. L &: R. J. Coleman. 18. Relating He-liothu spp. pheromone trap captUJes to egg countsin cotton: second year data &om dle delta of Mis.sissippi, pp. 313-317. In Proceedings. BeltwideCotton Production Research Conference. 2-7 Jan.1989. Nashville. TN. National Cotton Council ofAmerica. Memphis. TN.

Hayes, J. L., R. J. Coleman &: E. G. K:iD&o 1-. Re-lating Heliothu spp. pheromone b'ap captUJes toegg counts in cotton: data from dle delta of Missis-sippi. pp. ~268. In Proceedings. Beltwide Cot-ton Production Research Conference. 3-6 Jan.

A FREQUENT CRITmopathogenic hyplticides is the timetact with the infectinsect. During thifected pest continuto crops. Despite itstages, Beauveri6 twas shown to beColorado potato beeata (Say), in mals (

conditions (Fargue1981, Campbell e1988). Because oflast instar to fun!Schaerffenberg 1951960; Bajan &: KnSikura &: Sikura 16riod may induce seson, we investigafourth.instar Colo!with B. bassiana ir

Recei1>ed for publication 27 la_", 1993; accepted

24A~1993.

I Station de Recherc~tional de 1. Recherche Ianrourt Cedex, France.

2 Universit~ Paris VIIture, 2 Place Jussieu F .

3 Department of PlantRhode Island, Kinpton.