Discrimination by Ovipositing Boll Weevils(Coleoptera Curculionidae) Against Previously Infested

Hampea (Malvaceae) Flower Buds

PHILIP A STANSLY ANDJAMES R CATE

Department of Entomology Texas AampM UniversityCollege Station Texas 77843

Environ Enlomol 13 1361-1365 (1984)ABSTRACT The low occurrence of multiple oviposition punctures in flower buds of Ham-pea nutricia indicated an avoidance of already infested buds by female boll weevils Thisbehavior was confirmed by observations which showed that individual weevils inspectedbuds before oviposition and rejected them if a puncture plugged with frass was encounteredFree-choice tests showed that females preferred uninfested buds for oviposition over eitherinfested buds or buds implanted with a frass plug taken from an infested bud Forcingweevils to oviposit in infested buds greatly decreased egg production

EARLY IN the century it was observed that unin-fested cotton flower buds were selected by bollweevils for oviposition in preference to infestedbuds and generally only one egg was depositedper bud except in no-choice situations (Hunter andPierce 1912) In spite of this interesting observa-tion little subsequent work has been done to de-termine the mechanism underlying this behaviorAmong the studies conducted to date conclusionshave been contradictory Everett and Earle (1964)suggested that secretions possibly originating fromthe collateral gland are used by the weevil to sealoviposition punctures and could function to deterother females from ovipositing In a similar veinHedin et al (1974) mention a suggestion made byW H Cross that marking pheromone(s) which actas oviposition deterrents might occur in puncture-sealing frass Mitchell and Cross (1969) describedoviposition behavior of the boll weevil includinginspection of flower buds and an apparent abilityto perceive previous punctures Jenkins et al (1975)later cast doubt on the existence of such a deter-rent based on the frequency distribution of eggpunctures on cotton buds from fields in MississippiMcKibben et al (1982) attempted to resolve thiscontradiction by hypothesizing that exposure toinfested buds raises the response threshold Theseauthors showed that a computer simulation of thismodel generated a distribution of oviposit ionswhich was indistinguishable from random exceptat high infestation levels

In the face of the uncertainty concerning thisweevils ability to respond to infested buds itseemed useful to consider oviposition behavior ina population of boll weevils assumed to be isolatedfrom and unaffected by any changes that haveoccurred in the last 80 years in the weevils of theUnited States One such population occurs insoutheastern Mexico where the host plant is anabundant and fast-growing native tree Hampea

nutricia Fryxell (Malvaceae) There we carried outthe observations and experiments described belowwith the object of answering two related questions(1) Is the distribution of oviposition puncturesfound in the field independent of previous ovipo-sitions or does it indicate discrimination by wee-vils against punctured buds and (2) Can discrim-ination behavior be demonstrated with individualweevils under laboratory conditions Preliminaryresults were strongly in favor of the discriminationhypothesis and we were encouraged in our effortsto determine what stimulated this behavior

Materials and Methods

Field data on the distribution of ovipositionpunctures was limited to buds of staminate H nu-tricia trees Weevils do not ordinarily oviposit inbuds or fruit of pistillate trees of this dioeciousspecies (P Stansly unpublished data) Data werecollected from 18 August to 16 November 1980 atthree sites each ca 20 km in different directionsfrom Cardenas Tabasco Mexico Sixteen treesfrom 3 to 8 m in height were sampled at weeklyintervals by cutting branches at random withpruning shears or a tree-trimmer Buds from cutbranches were separated in the laboratory andscored for weevil punctures Damaged plant ma-terial was dissected to detect whether or not im-mature weevils were present

Preference tests and behavioral observationswere carried out in a laboratory at the ColegioSuperior de Agricultura Tropical Weevils werecollected in the field as adults (n = 3) or as larvaeand reared to adults in the laboratory (n = 16)Adults were maintained separately on a 13LllDphotoperiod at an average of 304DC (SD = 16range 255-341 DC) and 80 relative humidity(SD = 63 range 65-92) as measured by a me-chanical hydrothermograph Each weevil was fed

1361

1362 ENVIRONMENTAL ENTOMOLOGY Vol 13 no 5

daily by placing freshly cut branches of Hampeacontaining undamaged buds in 500-ml Erlenmey-er flasks filled with water and covered with a sleeveof nylon marquisette 10 cm in diameter and 60cm long The sleeve was taped to the bottom ofthe flask and tied at the top above the branch Eggproduction was recorded daily Choice tests lasting24 h were performed in the same cages by substi-tuting branches containing buds of predeterminednumber and type for the normal food ration In-fested buds were produced when needed by usingbranches that had been exposed to another weevilthe previous day adjusting the number of punc-tured and unpunctured buds by removal andmarking both old punctures and fresh buds withindia ink The ratio of uninfested to infested ma-terial ranged from 015 to 57 in these experimentsin order to include the infestation level at whichsome weevils would be forced either to oviposit ininfested buds or decrease their normal productionIn all other choice tests the number of treated andcontrol buds was kept approximately equal Num-ber of weevils used and the tests performed foreach comparison are shown in Table 1

In additional treatments frass plugs from in-fested buds obtained as described above were re-moved with jewelers forceps and implanted in asmall hole made in the corolla of a fresh bud whereoviposition normally occurs To test solubility ofthe active ingredients in a nonpolar solvent from25 to 100 frass plugs were removed from budsexposed to weevils the day before and soaked inpetroleum ether for from 1 to 5 days before im-plantation The extract so obtained was dripped onto fresh buds with a Pasteur pipette Activity ofthe treated plugs was also tested Depending onthe treatment controls consisted of clean buds budsimplanted with fresh plugs buds punched with asharp tool or buds treated with plain petroleumether

After each 24-h testing period branches weretaken from the cages and the buds removed andscored according to whether or not new puncturesand plugs had appeared Initially all newly punc-tured buds were dissected to verify the presenceof an egg However as the number of buds to bedissected increased and it was seen that the pres-ence of a plug almost invariably indicated thepresence of an egg (eggs per plug = 098 SE =0001 n = 62) only spot dissections were per-formed The use of sealed punctures was also con-sidered by Everett and Ray (1962) to be a reliableindication of oviposition by weevils in cotton Whenweevils were forced to oviposit in infested buds94 (SE = 04 n = 22) of the most recently madeplugged punctures contained eggs

Observations on ovipositional behavior weremade on three weevils by lowering the marqui-sette sleeve so that the weevils movements couldbe seen A stopwatch was used to time searchingand oviposition

Results

Field Observations A total of 3614 male Ham-pea buds were examined in 1980 Of these 964had one plugged puncture 7 had two and nonehad more than two If oviposition occurred inde-pendently of the presence or absence of a previouspuncture a Poisson model should fit these datawithout significant deviation However the actualdistribution varied significantly from that predict-ed by the Poisson (x2 = 1659 P lt 00001) (Sokaland Rohlf 1981) The largest deviation from thePoisson model occurred in the class of multiple-plugged punctures (expected = 109) The weevilsappeared to be avoiding ovipositing in infestedbuds providing clear evidence to reject the nullhypothesis of nondiscrimination

There were too few doubly punctured buds todiscern any particular pattern in their occurrenceThey were found over the entire season and therewas little correlation with date (r = 0039) or per-centage of buds punctured (r = 012) Two of thedoubly punctured buds contained no weevils andone contained a second-instar larva and an emptycell The four remaining buds all contained twoimmature weevils Two contained either eggs ornewly hatched first-instar larvae one had a secondinstar and a dead first-instar larva and one con-tained a second-instar and a small third-instar lar-va Judging from past experience all these larvaehad been developing for ca 1 or 2 days sincehatching

Behavioral Observations The ovipositing wee-vil pierced the corolla and androecium with itsmandibles (average time 96 min SE = 045 n =12) after which it made a 180middot rotation insertedthe ovipositor into the newly excavated hole anddeposited an egg (average time 13 min SE = 002n = 12) The female then sealed the puncture bydefecating into it tamping down the frass withseveral rapid vertical motions of the abdomen (Fig1) This tamped frass formed a plug approximately1 mm in diameter which was yellow at first (likethe Hampea pollen it was seen to contain) anddarkened to a reddish brown in a few days Vir-tually all oviposition punctures made by weevilsin the laboratory were sealed in this way (996SE = 02 n = 246)

The weevil rested for a period of 175 min(SE = 12 n = 8) usually at the base of the pedun-cle before resuming activity with a rapid searchalong the branch until it encountered anotherflower bud It then ran over the bud antennatingpaying special attention to the corolla On unin-fested buds average inspection time was 295 s (SE= 12 n = 8) before a site was selected and thepuncture begun However on infested buds (n =16) the frass plug was soon encountered anten-nated and even nibbled after which the weevilmoved off the bud to continue searching elsewhere(Fig 2)

October 1984 STAN$LY AND CATE BOLL WEEVILS ON Hampea 1363

Fig 1 Female boll weevil having just finished ovi-positing plugging puncture with frass

Laboratory Experiments In laboratory studiesthere was a highly significant preference for un-infested buds compared to infested buds for ovi-position (Xmiddot == 508 P lt 00001 Table la) Theseresults were highly significant despite the fact thatlow numbers of uninfested buds in some of thetrials forced weevils to oviposit in infested buds

Clean buds implanted with fresh ovipositionplugs were strongly avoided also (Xmiddot == 122 P lt00001 Table Ib) showing that the oviposition plugwas sufficient to initiate discriminatory behaviorRelatively fewer mistakes were made by theweevils in experiments with implanted buds thanin experiments with infested buds because in theformer there was always a choice

The following results are statistically significantthough possibly not definitive due to the smallnumber of animals tested Untreated control budswere preferred over buds implanted with plugssoaked in petroleum ether but these in turn werepreferred over buds implanted with fresh plugs(X2 == 301 P lt 00001 Xmiddot == 106 P lt 0001 Ta-bles 1 c and d) Thus some deterrent activity wasremoved by the ether treatment Treatment of budswith a petroleum ether extract of plugs did notaffect choice compared to controls treated withether alone (X2 == 032 P lt 057 Table Ie) There-fore deterrent substances removed by ether wereeither transformed in some way or were too diluteto cause discrimination

The data in Table l(f-j) demonstrate that apuncture alone did not elicit discriminatory be-havior but that deplugging infested buds did notcompletely remove the avoidance stimulus Budspunctured with a sharp tool were accepted asreadily as unpunctured buds (X2 == 0001 P == 097Table If) and were preferred both over puncturedbuds implanted with a plug (X == 370 P lt 00001

Fig 2 Searching female boll weevil homing in onfrass plug

Table Ig) and over infested buds with plugs re-moved (X == 501 P lt 00001 X == 155 P lt00001 Tables 1 h and i) When offered a choicebetween buds containing a plug and infested budswith the plug removed the expected frequencieswere too small to use the X test and a G test wasused instead (Sokal and Rohlf 1981) It showed nosignificant difference between deplugged infestedbuds and plugged infested buds as oviposition sites(G == 24 P lt 012 Table Ij) It is not clear if thisresult was due to the presence of active substancesseparate from the frass plug or to contaminationwith plug material itself

Virgin females (which seldom oviposited andusually on the outside of the bud) discriminatedagainst infested buds in their feeding but malesdid not (x2 == 398 P lt 00001 X2 = 22 P == 014Table 2) Thus only females discriminated againstinfested buds and they did so whether or not theywere ovipositing

There were data that indicated that the pres-ence of infested buds had a suppressive effect onoviposition Reproductively active weevils whenprovided only with un infested buds oviposited anaverage of 188 eggs per day (n == 85 SE == 11)In contrast when infested material was also pres-ent only 129 eggs per day (n == 63 SE == 08) werelaid which was significantly fewer (P lt 0001Wilcoxon paired ranks test) When the ratio ofinfested to un infested was 21 or greater these sameweevils oviposited only 102 eggs per day (n == 19SE == 11) Weevils forced to oviposit in infestedbuds by the lack of clean material decreased eggproduction to half or less of former values in somecases even stopping completely

Ten doubly infested buds were dissected 8 daysafter the second oviposition All 10 buds contained

1364 ENVIRONME~TAL ENTOMOLOGY Vol 13 no 5

Table I Numbers of H nufricia Hower buds accepted and not accepted by female boll weevils in oviposition choicetests data arranged in contingency tables and analyzed for independence by x2

la Accepted Not lb Accepted Notaccepted accepted

Uninfested Obs 649 500 Uninfested Obs 119 18Exp 384 764 Exp 84 53

Infested Obs 183 1155 Plug implanted Obs 3 60Exp 448 890 Exp 38 25

Number of weevils = 9 Number of weevils = 2Number of trials = 76 Number of trials = 5x2 = 508 P lt 00001 X2 = 122 P lt 00001

Ie Accepted Not Id Accepted Notaccepted accepted

Uninfested Obs 56 13 Fresh plug Obs 1 13Exp 42 27 Exp 6 20

Ether-soaked plug Obs 13 31 Obs 13 19Exp 7 11 Ether-soaked plug Exp 8 24

Number of weevils = 2 Number of weevils = 1Number of trials = 3 Number of trials = 2X2 = 301 P lt 00001 X2 = 106 P = 0001

Ie Accepted Not If Accepted Notaccepted accepted

Plug extract Obs 15 14 Whole buds Obs 39 8Exp 14 15 Exp 39 8

Ether Obs 17 21 Punched buds Obs 20 4Exp 18 20 Exp 20 4

Number of weevils = 1 Number of weevils = 2Number of trials = 1 Number of trials = 2X2 = 032 P = 057 X2 = 0001 P = 097

Ig Accepted Not 1h Accepted Notaccepted accepted

Plug implanted Obs 0 27 Uninfested Obs 28 1Exp 11 16 Exp 14 15

Bud punched Obs 20 4 Plug removed Obs 3 33Exp 9 15 Exp 17 19

Number of weevils = 2 Number of weevils = 2Number of trials = 2 Number of trials = 2X2 = 370 P lt 00001 X2 = 501 P lt 00001

Ii Accepted NotIj Accepted Not

accepted accepted

Bud punched Obs 12 1 Plug present Obs 0 36Exp 7 6 Exp 15 345

Plug removed Obs 3 13 Plug removed Obs 3 33Exp 8 8 Exp 15 345

Number of weevils = 1 Number of weevils = 2Number of trials = 1 Number of trials = 2X2 = 155 P lt 00001 G = 24 P=012

one weevil either a large third instar or a pupaIt must be assumed that one of the original wee-vils (probably the one younger by a day) waskilled and possibly consumed by the other

Discussion

Both field and laboratory data summarized aboveleave little doubt that boll weevils on Hampea inTabasco discriminate strongly against infested budsand that this behavior is responsible for the lowerobserved number of multiple ovipositions thanpredicted by a Poisson model The extreme scar-

city of multiply punctured Hampea buds in Ta-basco is in sharp contrast to the situation in Mis-sissippi cotton as reported by Jenkins et al (1975)These authors found 40 and 49 multiply punc-tured buds in the respective 2 years of their studyHow can this difference between two weevil pop-ulations considered conspecific (H R Burke per-sonal communication) be accounted for In theTabasco system the small size of Hampea buds isprobably the reason that no more than one weevilsurvives in cases of multiple oviposition The ul-timate death of all but the first egg laid must createstrong selective pressure to discriminate against an

October 1984 STANSLY AND CATE BOLL WEEVILS ON Hampea 1365

Table 2 Numbers of H nurricia flower buds accepted and not accepted for feeding by male (a) and virgin female(b) A grandi8 data arranged in contingency tables and analyzed for independence by X2

2a Accepted Notaccepted 2b Accepted Not

accepted

Uninfested

Infested

Obs 14 10Exp 12 14Obs 11 18Exp 15 17

Number of males = 3Number of trials = 2x2 = 22 P = 014

Uninfested

Infested

Obs 31 9Exp 21 32Obs 11 57Exp 35 55

Number of virgin females = 5Number of trials = 6X2= 398 P lt 00001

infested bud Hunter and Pierce (1912) noted thatonly one weevil emerged from a cotton bud re-gardless of the number of ovipositions In morerecent times it is known that some second ovipo-sit ions are successful particularly in cotton fruit(bolls) and large buds (Hunter et al 1965) It iscommon for numerous weevils to emerge from aboll which may become considerably larger thanthe bud and still be susceptible to weevils (Hunterand Pierce 1912) whereas Hampea fruit is neversuccessfully attacked In cotton fields buds arescarce early in the season while in late seasonthere are few un infested buds or bolls At suchtimes the advantages of multiple oviposition overno oviposition would select for a less discriminat-ing weevil Such selection could act by raising thethreshold at which the marking pheromone is per-ceived by or responded to by the weevil Such amechanism can explain the differences in discrim-inatory behavior observed in cotton and HampeaIf the reported contrast between modern cottonweevils and those studied at the turn of the cen-tury is real a modern evolutionary trend towardnondiscrimination may be indicated Even todaydifferences in degree of discriminatory behaviorhave been observed between weevil strains on cot-ton (J N Jenkins personal communication) Suchdifferences may explain some of the discrepanciesreported It should be noted that it is not knownwhether the weakness or lack of discriminatorybehavior in these modern cotton weevils is due tothe absence of stimulatory substances or the in-ability to respond to such substances under certainconditions as suggested by McKibben et al (1982)

It is also of more than academic interest to de-termine the biological origin and chemical natureof the deterrent Virtually nothing is known aboutit except for its moderate insolubility in petro-leum ether The putative substance(s) could eitherbe altered or unaltered constituents of Hampeapollen or synthesized de novo by the weevil Iso-lation and identification of the active materialsfrom frass plugs themselves would be an importantstep in gaining a better understanding of weevilbiology and might also provide a powerful re-search and management tool

Acknowledgment

We thank the faculty and staff of the Colegio Superiorde Agricultura Tropical for use of facilities and for theirhelp and cooperation during the course of this and otherstudies conducted in Tabasco The following is a partiallist of persons whose remarks and criticisms we appre-ciated H R Burke A P Bhaktar O C Jenkins S JJohnston P C Krauter B F Vaissiere J K Walkerand the late W H Cross This material is based uponwork supported by USDA under Agreements No 71-59-2481-1-2-039-1 No 82 CSRS-2-1000 and No 58-519B-I-996-55290-15-4290 Approved as TA 19456 bythe Director Texas Agric Exp Stn

References Cited

Everett T H and N W Earle 1964 Boll weeviloviposition response in cotton squares and variousother substrates J Econ Entomol 57 561-566

Everett T H and J O Hay 1962 Observations onpuncturing and oviposition behavior of boll weevilIbid 55 634-642

Hedin P A H C Gueldner H D Henson and A CThompson 1974 Volatile constituents of male andfemale boll weevils and their frass J Insect Physiol20 2135-2142

Hunter H C T F Leigh C Lincoln B A Waddleand L A Bariola 1965 Evaluation of a selectedcross-section of cottons for resistance to the boll wee-vil Ark Agric Exp Stn Bull 700

Hunter W D and W D Pierce 1912 The Mexi-can boll weevil a summary of the investigations ofthis insect up to Dec 31 1911 US Dep Agric BurEntomol Bull 114

Jenkins J J W L Parrott and J W Jones 1975Boll weevil oviposition behavior multiple puncturedsquares Environ Entomol 4 861-867

McKibben G W L McGovern and W A Dickerson1982 Boll weevil (Coleoptera Curculionidae) ovi-position behavior a simulation analysis J EconEntomol 75 928-931

Mitchell H C and W H Cross 1969 Ovipositionby the boll weevil in the field Ibid 62 604-605

Sokal H H and F J Hohlf 1981 Biometry theprinciples and practice of statistics in biological re-search W H Freeman San Francisco

Received for publication 16 March 1984 accepted 5June 1984

1362 ENVIRONMENTAL ENTOMOLOGY Vol 13 no 5

daily by placing freshly cut branches of Hampeacontaining undamaged buds in 500-ml Erlenmey-er flasks filled with water and covered with a sleeveof nylon marquisette 10 cm in diameter and 60cm long The sleeve was taped to the bottom ofthe flask and tied at the top above the branch Eggproduction was recorded daily Choice tests lasting24 h were performed in the same cages by substi-tuting branches containing buds of predeterminednumber and type for the normal food ration In-fested buds were produced when needed by usingbranches that had been exposed to another weevilthe previous day adjusting the number of punc-tured and unpunctured buds by removal andmarking both old punctures and fresh buds withindia ink The ratio of uninfested to infested ma-terial ranged from 015 to 57 in these experimentsin order to include the infestation level at whichsome weevils would be forced either to oviposit ininfested buds or decrease their normal productionIn all other choice tests the number of treated andcontrol buds was kept approximately equal Num-ber of weevils used and the tests performed foreach comparison are shown in Table 1

In additional treatments frass plugs from in-fested buds obtained as described above were re-moved with jewelers forceps and implanted in asmall hole made in the corolla of a fresh bud whereoviposition normally occurs To test solubility ofthe active ingredients in a nonpolar solvent from25 to 100 frass plugs were removed from budsexposed to weevils the day before and soaked inpetroleum ether for from 1 to 5 days before im-plantation The extract so obtained was dripped onto fresh buds with a Pasteur pipette Activity ofthe treated plugs was also tested Depending onthe treatment controls consisted of clean buds budsimplanted with fresh plugs buds punched with asharp tool or buds treated with plain petroleumether

After each 24-h testing period branches weretaken from the cages and the buds removed andscored according to whether or not new puncturesand plugs had appeared Initially all newly punc-tured buds were dissected to verify the presenceof an egg However as the number of buds to bedissected increased and it was seen that the pres-ence of a plug almost invariably indicated thepresence of an egg (eggs per plug = 098 SE =0001 n = 62) only spot dissections were per-formed The use of sealed punctures was also con-sidered by Everett and Ray (1962) to be a reliableindication of oviposition by weevils in cotton Whenweevils were forced to oviposit in infested buds94 (SE = 04 n = 22) of the most recently madeplugged punctures contained eggs

Observations on ovipositional behavior weremade on three weevils by lowering the marqui-sette sleeve so that the weevils movements couldbe seen A stopwatch was used to time searchingand oviposition

Results

Field Observations A total of 3614 male Ham-pea buds were examined in 1980 Of these 964had one plugged puncture 7 had two and nonehad more than two If oviposition occurred inde-pendently of the presence or absence of a previouspuncture a Poisson model should fit these datawithout significant deviation However the actualdistribution varied significantly from that predict-ed by the Poisson (x2 = 1659 P lt 00001) (Sokaland Rohlf 1981) The largest deviation from thePoisson model occurred in the class of multiple-plugged punctures (expected = 109) The weevilsappeared to be avoiding ovipositing in infestedbuds providing clear evidence to reject the nullhypothesis of nondiscrimination

There were too few doubly punctured buds todiscern any particular pattern in their occurrenceThey were found over the entire season and therewas little correlation with date (r = 0039) or per-centage of buds punctured (r = 012) Two of thedoubly punctured buds contained no weevils andone contained a second-instar larva and an emptycell The four remaining buds all contained twoimmature weevils Two contained either eggs ornewly hatched first-instar larvae one had a secondinstar and a dead first-instar larva and one con-tained a second-instar and a small third-instar lar-va Judging from past experience all these larvaehad been developing for ca 1 or 2 days sincehatching

Behavioral Observations The ovipositing wee-vil pierced the corolla and androecium with itsmandibles (average time 96 min SE = 045 n =12) after which it made a 180middot rotation insertedthe ovipositor into the newly excavated hole anddeposited an egg (average time 13 min SE = 002n = 12) The female then sealed the puncture bydefecating into it tamping down the frass withseveral rapid vertical motions of the abdomen (Fig1) This tamped frass formed a plug approximately1 mm in diameter which was yellow at first (likethe Hampea pollen it was seen to contain) anddarkened to a reddish brown in a few days Vir-tually all oviposition punctures made by weevilsin the laboratory were sealed in this way (996SE = 02 n = 246)

The weevil rested for a period of 175 min(SE = 12 n = 8) usually at the base of the pedun-cle before resuming activity with a rapid searchalong the branch until it encountered anotherflower bud It then ran over the bud antennatingpaying special attention to the corolla On unin-fested buds average inspection time was 295 s (SE= 12 n = 8) before a site was selected and thepuncture begun However on infested buds (n =16) the frass plug was soon encountered anten-nated and even nibbled after which the weevilmoved off the bud to continue searching elsewhere(Fig 2)

October 1984 STAN$LY AND CATE BOLL WEEVILS ON Hampea 1363

Fig 1 Female boll weevil having just finished ovi-positing plugging puncture with frass

Laboratory Experiments In laboratory studiesthere was a highly significant preference for un-infested buds compared to infested buds for ovi-position (Xmiddot == 508 P lt 00001 Table la) Theseresults were highly significant despite the fact thatlow numbers of uninfested buds in some of thetrials forced weevils to oviposit in infested buds

Clean buds implanted with fresh ovipositionplugs were strongly avoided also (Xmiddot == 122 P lt00001 Table Ib) showing that the oviposition plugwas sufficient to initiate discriminatory behaviorRelatively fewer mistakes were made by theweevils in experiments with implanted buds thanin experiments with infested buds because in theformer there was always a choice

The following results are statistically significantthough possibly not definitive due to the smallnumber of animals tested Untreated control budswere preferred over buds implanted with plugssoaked in petroleum ether but these in turn werepreferred over buds implanted with fresh plugs(X2 == 301 P lt 00001 Xmiddot == 106 P lt 0001 Ta-bles 1 c and d) Thus some deterrent activity wasremoved by the ether treatment Treatment of budswith a petroleum ether extract of plugs did notaffect choice compared to controls treated withether alone (X2 == 032 P lt 057 Table Ie) There-fore deterrent substances removed by ether wereeither transformed in some way or were too diluteto cause discrimination

The data in Table l(f-j) demonstrate that apuncture alone did not elicit discriminatory be-havior but that deplugging infested buds did notcompletely remove the avoidance stimulus Budspunctured with a sharp tool were accepted asreadily as unpunctured buds (X2 == 0001 P == 097Table If) and were preferred both over puncturedbuds implanted with a plug (X == 370 P lt 00001

Fig 2 Searching female boll weevil homing in onfrass plug

Table Ig) and over infested buds with plugs re-moved (X == 501 P lt 00001 X == 155 P lt00001 Tables 1 h and i) When offered a choicebetween buds containing a plug and infested budswith the plug removed the expected frequencieswere too small to use the X test and a G test wasused instead (Sokal and Rohlf 1981) It showed nosignificant difference between deplugged infestedbuds and plugged infested buds as oviposition sites(G == 24 P lt 012 Table Ij) It is not clear if thisresult was due to the presence of active substancesseparate from the frass plug or to contaminationwith plug material itself

Virgin females (which seldom oviposited andusually on the outside of the bud) discriminatedagainst infested buds in their feeding but malesdid not (x2 == 398 P lt 00001 X2 = 22 P == 014Table 2) Thus only females discriminated againstinfested buds and they did so whether or not theywere ovipositing

There were data that indicated that the pres-ence of infested buds had a suppressive effect onoviposition Reproductively active weevils whenprovided only with un infested buds oviposited anaverage of 188 eggs per day (n == 85 SE == 11)In contrast when infested material was also pres-ent only 129 eggs per day (n == 63 SE == 08) werelaid which was significantly fewer (P lt 0001Wilcoxon paired ranks test) When the ratio ofinfested to un infested was 21 or greater these sameweevils oviposited only 102 eggs per day (n == 19SE == 11) Weevils forced to oviposit in infestedbuds by the lack of clean material decreased eggproduction to half or less of former values in somecases even stopping completely

Ten doubly infested buds were dissected 8 daysafter the second oviposition All 10 buds contained

1364 ENVIRONME~TAL ENTOMOLOGY Vol 13 no 5

Table I Numbers of H nufricia Hower buds accepted and not accepted by female boll weevils in oviposition choicetests data arranged in contingency tables and analyzed for independence by x2

la Accepted Not lb Accepted Notaccepted accepted

Uninfested Obs 649 500 Uninfested Obs 119 18Exp 384 764 Exp 84 53

Infested Obs 183 1155 Plug implanted Obs 3 60Exp 448 890 Exp 38 25

Number of weevils = 9 Number of weevils = 2Number of trials = 76 Number of trials = 5x2 = 508 P lt 00001 X2 = 122 P lt 00001

Ie Accepted Not Id Accepted Notaccepted accepted

Uninfested Obs 56 13 Fresh plug Obs 1 13Exp 42 27 Exp 6 20

Ether-soaked plug Obs 13 31 Obs 13 19Exp 7 11 Ether-soaked plug Exp 8 24

Number of weevils = 2 Number of weevils = 1Number of trials = 3 Number of trials = 2X2 = 301 P lt 00001 X2 = 106 P = 0001

Ie Accepted Not If Accepted Notaccepted accepted

Plug extract Obs 15 14 Whole buds Obs 39 8Exp 14 15 Exp 39 8

Ether Obs 17 21 Punched buds Obs 20 4Exp 18 20 Exp 20 4

Number of weevils = 1 Number of weevils = 2Number of trials = 1 Number of trials = 2X2 = 032 P = 057 X2 = 0001 P = 097

Ig Accepted Not 1h Accepted Notaccepted accepted

Plug implanted Obs 0 27 Uninfested Obs 28 1Exp 11 16 Exp 14 15

Bud punched Obs 20 4 Plug removed Obs 3 33Exp 9 15 Exp 17 19

Number of weevils = 2 Number of weevils = 2Number of trials = 2 Number of trials = 2X2 = 370 P lt 00001 X2 = 501 P lt 00001

Ii Accepted NotIj Accepted Not

accepted accepted

Bud punched Obs 12 1 Plug present Obs 0 36Exp 7 6 Exp 15 345

Plug removed Obs 3 13 Plug removed Obs 3 33Exp 8 8 Exp 15 345

Number of weevils = 1 Number of weevils = 2Number of trials = 1 Number of trials = 2X2 = 155 P lt 00001 G = 24 P=012

one weevil either a large third instar or a pupaIt must be assumed that one of the original wee-vils (probably the one younger by a day) waskilled and possibly consumed by the other

Discussion

Both field and laboratory data summarized aboveleave little doubt that boll weevils on Hampea inTabasco discriminate strongly against infested budsand that this behavior is responsible for the lowerobserved number of multiple ovipositions thanpredicted by a Poisson model The extreme scar-

city of multiply punctured Hampea buds in Ta-basco is in sharp contrast to the situation in Mis-sissippi cotton as reported by Jenkins et al (1975)These authors found 40 and 49 multiply punc-tured buds in the respective 2 years of their studyHow can this difference between two weevil pop-ulations considered conspecific (H R Burke per-sonal communication) be accounted for In theTabasco system the small size of Hampea buds isprobably the reason that no more than one weevilsurvives in cases of multiple oviposition The ul-timate death of all but the first egg laid must createstrong selective pressure to discriminate against an

October 1984 STANSLY AND CATE BOLL WEEVILS ON Hampea 1365

Table 2 Numbers of H nurricia flower buds accepted and not accepted for feeding by male (a) and virgin female(b) A grandi8 data arranged in contingency tables and analyzed for independence by X2

2a Accepted Notaccepted 2b Accepted Not

accepted

Uninfested

Infested

Obs 14 10Exp 12 14Obs 11 18Exp 15 17

Number of males = 3Number of trials = 2x2 = 22 P = 014

Uninfested

Infested

Obs 31 9Exp 21 32Obs 11 57Exp 35 55

Number of virgin females = 5Number of trials = 6X2= 398 P lt 00001

infested bud Hunter and Pierce (1912) noted thatonly one weevil emerged from a cotton bud re-gardless of the number of ovipositions In morerecent times it is known that some second ovipo-sit ions are successful particularly in cotton fruit(bolls) and large buds (Hunter et al 1965) It iscommon for numerous weevils to emerge from aboll which may become considerably larger thanthe bud and still be susceptible to weevils (Hunterand Pierce 1912) whereas Hampea fruit is neversuccessfully attacked In cotton fields buds arescarce early in the season while in late seasonthere are few un infested buds or bolls At suchtimes the advantages of multiple oviposition overno oviposition would select for a less discriminat-ing weevil Such selection could act by raising thethreshold at which the marking pheromone is per-ceived by or responded to by the weevil Such amechanism can explain the differences in discrim-inatory behavior observed in cotton and HampeaIf the reported contrast between modern cottonweevils and those studied at the turn of the cen-tury is real a modern evolutionary trend towardnondiscrimination may be indicated Even todaydifferences in degree of discriminatory behaviorhave been observed between weevil strains on cot-ton (J N Jenkins personal communication) Suchdifferences may explain some of the discrepanciesreported It should be noted that it is not knownwhether the weakness or lack of discriminatorybehavior in these modern cotton weevils is due tothe absence of stimulatory substances or the in-ability to respond to such substances under certainconditions as suggested by McKibben et al (1982)

It is also of more than academic interest to de-termine the biological origin and chemical natureof the deterrent Virtually nothing is known aboutit except for its moderate insolubility in petro-leum ether The putative substance(s) could eitherbe altered or unaltered constituents of Hampeapollen or synthesized de novo by the weevil Iso-lation and identification of the active materialsfrom frass plugs themselves would be an importantstep in gaining a better understanding of weevilbiology and might also provide a powerful re-search and management tool

Acknowledgment

We thank the faculty and staff of the Colegio Superiorde Agricultura Tropical for use of facilities and for theirhelp and cooperation during the course of this and otherstudies conducted in Tabasco The following is a partiallist of persons whose remarks and criticisms we appre-ciated H R Burke A P Bhaktar O C Jenkins S JJohnston P C Krauter B F Vaissiere J K Walkerand the late W H Cross This material is based uponwork supported by USDA under Agreements No 71-59-2481-1-2-039-1 No 82 CSRS-2-1000 and No 58-519B-I-996-55290-15-4290 Approved as TA 19456 bythe Director Texas Agric Exp Stn

References Cited

Everett T H and N W Earle 1964 Boll weeviloviposition response in cotton squares and variousother substrates J Econ Entomol 57 561-566

Everett T H and J O Hay 1962 Observations onpuncturing and oviposition behavior of boll weevilIbid 55 634-642

Hedin P A H C Gueldner H D Henson and A CThompson 1974 Volatile constituents of male andfemale boll weevils and their frass J Insect Physiol20 2135-2142

Hunter H C T F Leigh C Lincoln B A Waddleand L A Bariola 1965 Evaluation of a selectedcross-section of cottons for resistance to the boll wee-vil Ark Agric Exp Stn Bull 700

Hunter W D and W D Pierce 1912 The Mexi-can boll weevil a summary of the investigations ofthis insect up to Dec 31 1911 US Dep Agric BurEntomol Bull 114

Jenkins J J W L Parrott and J W Jones 1975Boll weevil oviposition behavior multiple puncturedsquares Environ Entomol 4 861-867

McKibben G W L McGovern and W A Dickerson1982 Boll weevil (Coleoptera Curculionidae) ovi-position behavior a simulation analysis J EconEntomol 75 928-931

Mitchell H C and W H Cross 1969 Ovipositionby the boll weevil in the field Ibid 62 604-605

Sokal H H and F J Hohlf 1981 Biometry theprinciples and practice of statistics in biological re-search W H Freeman San Francisco

Received for publication 16 March 1984 accepted 5June 1984

October 1984 STAN$LY AND CATE BOLL WEEVILS ON Hampea 1363

Fig 1 Female boll weevil having just finished ovi-positing plugging puncture with frass

Laboratory Experiments In laboratory studiesthere was a highly significant preference for un-infested buds compared to infested buds for ovi-position (Xmiddot == 508 P lt 00001 Table la) Theseresults were highly significant despite the fact thatlow numbers of uninfested buds in some of thetrials forced weevils to oviposit in infested buds

Clean buds implanted with fresh ovipositionplugs were strongly avoided also (Xmiddot == 122 P lt00001 Table Ib) showing that the oviposition plugwas sufficient to initiate discriminatory behaviorRelatively fewer mistakes were made by theweevils in experiments with implanted buds thanin experiments with infested buds because in theformer there was always a choice

The following results are statistically significantthough possibly not definitive due to the smallnumber of animals tested Untreated control budswere preferred over buds implanted with plugssoaked in petroleum ether but these in turn werepreferred over buds implanted with fresh plugs(X2 == 301 P lt 00001 Xmiddot == 106 P lt 0001 Ta-bles 1 c and d) Thus some deterrent activity wasremoved by the ether treatment Treatment of budswith a petroleum ether extract of plugs did notaffect choice compared to controls treated withether alone (X2 == 032 P lt 057 Table Ie) There-fore deterrent substances removed by ether wereeither transformed in some way or were too diluteto cause discrimination

The data in Table l(f-j) demonstrate that apuncture alone did not elicit discriminatory be-havior but that deplugging infested buds did notcompletely remove the avoidance stimulus Budspunctured with a sharp tool were accepted asreadily as unpunctured buds (X2 == 0001 P == 097Table If) and were preferred both over puncturedbuds implanted with a plug (X == 370 P lt 00001

Fig 2 Searching female boll weevil homing in onfrass plug

Table Ig) and over infested buds with plugs re-moved (X == 501 P lt 00001 X == 155 P lt00001 Tables 1 h and i) When offered a choicebetween buds containing a plug and infested budswith the plug removed the expected frequencieswere too small to use the X test and a G test wasused instead (Sokal and Rohlf 1981) It showed nosignificant difference between deplugged infestedbuds and plugged infested buds as oviposition sites(G == 24 P lt 012 Table Ij) It is not clear if thisresult was due to the presence of active substancesseparate from the frass plug or to contaminationwith plug material itself

Virgin females (which seldom oviposited andusually on the outside of the bud) discriminatedagainst infested buds in their feeding but malesdid not (x2 == 398 P lt 00001 X2 = 22 P == 014Table 2) Thus only females discriminated againstinfested buds and they did so whether or not theywere ovipositing

There were data that indicated that the pres-ence of infested buds had a suppressive effect onoviposition Reproductively active weevils whenprovided only with un infested buds oviposited anaverage of 188 eggs per day (n == 85 SE == 11)In contrast when infested material was also pres-ent only 129 eggs per day (n == 63 SE == 08) werelaid which was significantly fewer (P lt 0001Wilcoxon paired ranks test) When the ratio ofinfested to un infested was 21 or greater these sameweevils oviposited only 102 eggs per day (n == 19SE == 11) Weevils forced to oviposit in infestedbuds by the lack of clean material decreased eggproduction to half or less of former values in somecases even stopping completely

Ten doubly infested buds were dissected 8 daysafter the second oviposition All 10 buds contained

1364 ENVIRONME~TAL ENTOMOLOGY Vol 13 no 5

Table I Numbers of H nufricia Hower buds accepted and not accepted by female boll weevils in oviposition choicetests data arranged in contingency tables and analyzed for independence by x2

la Accepted Not lb Accepted Notaccepted accepted

Uninfested Obs 649 500 Uninfested Obs 119 18Exp 384 764 Exp 84 53

Infested Obs 183 1155 Plug implanted Obs 3 60Exp 448 890 Exp 38 25

Number of weevils = 9 Number of weevils = 2Number of trials = 76 Number of trials = 5x2 = 508 P lt 00001 X2 = 122 P lt 00001

Ie Accepted Not Id Accepted Notaccepted accepted

Uninfested Obs 56 13 Fresh plug Obs 1 13Exp 42 27 Exp 6 20

Ether-soaked plug Obs 13 31 Obs 13 19Exp 7 11 Ether-soaked plug Exp 8 24

Number of weevils = 2 Number of weevils = 1Number of trials = 3 Number of trials = 2X2 = 301 P lt 00001 X2 = 106 P = 0001

Ie Accepted Not If Accepted Notaccepted accepted

Plug extract Obs 15 14 Whole buds Obs 39 8Exp 14 15 Exp 39 8

Ether Obs 17 21 Punched buds Obs 20 4Exp 18 20 Exp 20 4

Number of weevils = 1 Number of weevils = 2Number of trials = 1 Number of trials = 2X2 = 032 P = 057 X2 = 0001 P = 097

Ig Accepted Not 1h Accepted Notaccepted accepted

Plug implanted Obs 0 27 Uninfested Obs 28 1Exp 11 16 Exp 14 15

Bud punched Obs 20 4 Plug removed Obs 3 33Exp 9 15 Exp 17 19

Number of weevils = 2 Number of weevils = 2Number of trials = 2 Number of trials = 2X2 = 370 P lt 00001 X2 = 501 P lt 00001

Ii Accepted NotIj Accepted Not

accepted accepted

Bud punched Obs 12 1 Plug present Obs 0 36Exp 7 6 Exp 15 345

Plug removed Obs 3 13 Plug removed Obs 3 33Exp 8 8 Exp 15 345

Number of weevils = 1 Number of weevils = 2Number of trials = 1 Number of trials = 2X2 = 155 P lt 00001 G = 24 P=012

one weevil either a large third instar or a pupaIt must be assumed that one of the original wee-vils (probably the one younger by a day) waskilled and possibly consumed by the other

Discussion

Both field and laboratory data summarized aboveleave little doubt that boll weevils on Hampea inTabasco discriminate strongly against infested budsand that this behavior is responsible for the lowerobserved number of multiple ovipositions thanpredicted by a Poisson model The extreme scar-

city of multiply punctured Hampea buds in Ta-basco is in sharp contrast to the situation in Mis-sissippi cotton as reported by Jenkins et al (1975)These authors found 40 and 49 multiply punc-tured buds in the respective 2 years of their studyHow can this difference between two weevil pop-ulations considered conspecific (H R Burke per-sonal communication) be accounted for In theTabasco system the small size of Hampea buds isprobably the reason that no more than one weevilsurvives in cases of multiple oviposition The ul-timate death of all but the first egg laid must createstrong selective pressure to discriminate against an

October 1984 STANSLY AND CATE BOLL WEEVILS ON Hampea 1365

Table 2 Numbers of H nurricia flower buds accepted and not accepted for feeding by male (a) and virgin female(b) A grandi8 data arranged in contingency tables and analyzed for independence by X2

2a Accepted Notaccepted 2b Accepted Not

accepted

Uninfested

Infested

Obs 14 10Exp 12 14Obs 11 18Exp 15 17

Number of males = 3Number of trials = 2x2 = 22 P = 014

Uninfested

Infested

Obs 31 9Exp 21 32Obs 11 57Exp 35 55

Number of virgin females = 5Number of trials = 6X2= 398 P lt 00001

infested bud Hunter and Pierce (1912) noted thatonly one weevil emerged from a cotton bud re-gardless of the number of ovipositions In morerecent times it is known that some second ovipo-sit ions are successful particularly in cotton fruit(bolls) and large buds (Hunter et al 1965) It iscommon for numerous weevils to emerge from aboll which may become considerably larger thanthe bud and still be susceptible to weevils (Hunterand Pierce 1912) whereas Hampea fruit is neversuccessfully attacked In cotton fields buds arescarce early in the season while in late seasonthere are few un infested buds or bolls At suchtimes the advantages of multiple oviposition overno oviposition would select for a less discriminat-ing weevil Such selection could act by raising thethreshold at which the marking pheromone is per-ceived by or responded to by the weevil Such amechanism can explain the differences in discrim-inatory behavior observed in cotton and HampeaIf the reported contrast between modern cottonweevils and those studied at the turn of the cen-tury is real a modern evolutionary trend towardnondiscrimination may be indicated Even todaydifferences in degree of discriminatory behaviorhave been observed between weevil strains on cot-ton (J N Jenkins personal communication) Suchdifferences may explain some of the discrepanciesreported It should be noted that it is not knownwhether the weakness or lack of discriminatorybehavior in these modern cotton weevils is due tothe absence of stimulatory substances or the in-ability to respond to such substances under certainconditions as suggested by McKibben et al (1982)

It is also of more than academic interest to de-termine the biological origin and chemical natureof the deterrent Virtually nothing is known aboutit except for its moderate insolubility in petro-leum ether The putative substance(s) could eitherbe altered or unaltered constituents of Hampeapollen or synthesized de novo by the weevil Iso-lation and identification of the active materialsfrom frass plugs themselves would be an importantstep in gaining a better understanding of weevilbiology and might also provide a powerful re-search and management tool

Acknowledgment

We thank the faculty and staff of the Colegio Superiorde Agricultura Tropical for use of facilities and for theirhelp and cooperation during the course of this and otherstudies conducted in Tabasco The following is a partiallist of persons whose remarks and criticisms we appre-ciated H R Burke A P Bhaktar O C Jenkins S JJohnston P C Krauter B F Vaissiere J K Walkerand the late W H Cross This material is based uponwork supported by USDA under Agreements No 71-59-2481-1-2-039-1 No 82 CSRS-2-1000 and No 58-519B-I-996-55290-15-4290 Approved as TA 19456 bythe Director Texas Agric Exp Stn

References Cited

Everett T H and N W Earle 1964 Boll weeviloviposition response in cotton squares and variousother substrates J Econ Entomol 57 561-566

Everett T H and J O Hay 1962 Observations onpuncturing and oviposition behavior of boll weevilIbid 55 634-642

Hedin P A H C Gueldner H D Henson and A CThompson 1974 Volatile constituents of male andfemale boll weevils and their frass J Insect Physiol20 2135-2142

Hunter H C T F Leigh C Lincoln B A Waddleand L A Bariola 1965 Evaluation of a selectedcross-section of cottons for resistance to the boll wee-vil Ark Agric Exp Stn Bull 700

Hunter W D and W D Pierce 1912 The Mexi-can boll weevil a summary of the investigations ofthis insect up to Dec 31 1911 US Dep Agric BurEntomol Bull 114

Jenkins J J W L Parrott and J W Jones 1975Boll weevil oviposition behavior multiple puncturedsquares Environ Entomol 4 861-867

McKibben G W L McGovern and W A Dickerson1982 Boll weevil (Coleoptera Curculionidae) ovi-position behavior a simulation analysis J EconEntomol 75 928-931

Mitchell H C and W H Cross 1969 Ovipositionby the boll weevil in the field Ibid 62 604-605

Sokal H H and F J Hohlf 1981 Biometry theprinciples and practice of statistics in biological re-search W H Freeman San Francisco

Received for publication 16 March 1984 accepted 5June 1984

1364 ENVIRONME~TAL ENTOMOLOGY Vol 13 no 5

Table I Numbers of H nufricia Hower buds accepted and not accepted by female boll weevils in oviposition choicetests data arranged in contingency tables and analyzed for independence by x2

la Accepted Not lb Accepted Notaccepted accepted

Uninfested Obs 649 500 Uninfested Obs 119 18Exp 384 764 Exp 84 53

Infested Obs 183 1155 Plug implanted Obs 3 60Exp 448 890 Exp 38 25

Number of weevils = 9 Number of weevils = 2Number of trials = 76 Number of trials = 5x2 = 508 P lt 00001 X2 = 122 P lt 00001

Ie Accepted Not Id Accepted Notaccepted accepted

Uninfested Obs 56 13 Fresh plug Obs 1 13Exp 42 27 Exp 6 20

Ether-soaked plug Obs 13 31 Obs 13 19Exp 7 11 Ether-soaked plug Exp 8 24

Number of weevils = 2 Number of weevils = 1Number of trials = 3 Number of trials = 2X2 = 301 P lt 00001 X2 = 106 P = 0001

Ie Accepted Not If Accepted Notaccepted accepted

Plug extract Obs 15 14 Whole buds Obs 39 8Exp 14 15 Exp 39 8

Ether Obs 17 21 Punched buds Obs 20 4Exp 18 20 Exp 20 4

Number of weevils = 1 Number of weevils = 2Number of trials = 1 Number of trials = 2X2 = 032 P = 057 X2 = 0001 P = 097

Ig Accepted Not 1h Accepted Notaccepted accepted

Plug implanted Obs 0 27 Uninfested Obs 28 1Exp 11 16 Exp 14 15

Bud punched Obs 20 4 Plug removed Obs 3 33Exp 9 15 Exp 17 19

Number of weevils = 2 Number of weevils = 2Number of trials = 2 Number of trials = 2X2 = 370 P lt 00001 X2 = 501 P lt 00001

Ii Accepted NotIj Accepted Not

accepted accepted

Bud punched Obs 12 1 Plug present Obs 0 36Exp 7 6 Exp 15 345

Plug removed Obs 3 13 Plug removed Obs 3 33Exp 8 8 Exp 15 345

Number of weevils = 1 Number of weevils = 2Number of trials = 1 Number of trials = 2X2 = 155 P lt 00001 G = 24 P=012

one weevil either a large third instar or a pupaIt must be assumed that one of the original wee-vils (probably the one younger by a day) waskilled and possibly consumed by the other

Discussion

Both field and laboratory data summarized aboveleave little doubt that boll weevils on Hampea inTabasco discriminate strongly against infested budsand that this behavior is responsible for the lowerobserved number of multiple ovipositions thanpredicted by a Poisson model The extreme scar-

city of multiply punctured Hampea buds in Ta-basco is in sharp contrast to the situation in Mis-sissippi cotton as reported by Jenkins et al (1975)These authors found 40 and 49 multiply punc-tured buds in the respective 2 years of their studyHow can this difference between two weevil pop-ulations considered conspecific (H R Burke per-sonal communication) be accounted for In theTabasco system the small size of Hampea buds isprobably the reason that no more than one weevilsurvives in cases of multiple oviposition The ul-timate death of all but the first egg laid must createstrong selective pressure to discriminate against an

October 1984 STANSLY AND CATE BOLL WEEVILS ON Hampea 1365

Table 2 Numbers of H nurricia flower buds accepted and not accepted for feeding by male (a) and virgin female(b) A grandi8 data arranged in contingency tables and analyzed for independence by X2

2a Accepted Notaccepted 2b Accepted Not

accepted

Uninfested

Infested

Obs 14 10Exp 12 14Obs 11 18Exp 15 17

Number of males = 3Number of trials = 2x2 = 22 P = 014

Uninfested

Infested

Obs 31 9Exp 21 32Obs 11 57Exp 35 55

Number of virgin females = 5Number of trials = 6X2= 398 P lt 00001

infested bud Hunter and Pierce (1912) noted thatonly one weevil emerged from a cotton bud re-gardless of the number of ovipositions In morerecent times it is known that some second ovipo-sit ions are successful particularly in cotton fruit(bolls) and large buds (Hunter et al 1965) It iscommon for numerous weevils to emerge from aboll which may become considerably larger thanthe bud and still be susceptible to weevils (Hunterand Pierce 1912) whereas Hampea fruit is neversuccessfully attacked In cotton fields buds arescarce early in the season while in late seasonthere are few un infested buds or bolls At suchtimes the advantages of multiple oviposition overno oviposition would select for a less discriminat-ing weevil Such selection could act by raising thethreshold at which the marking pheromone is per-ceived by or responded to by the weevil Such amechanism can explain the differences in discrim-inatory behavior observed in cotton and HampeaIf the reported contrast between modern cottonweevils and those studied at the turn of the cen-tury is real a modern evolutionary trend towardnondiscrimination may be indicated Even todaydifferences in degree of discriminatory behaviorhave been observed between weevil strains on cot-ton (J N Jenkins personal communication) Suchdifferences may explain some of the discrepanciesreported It should be noted that it is not knownwhether the weakness or lack of discriminatorybehavior in these modern cotton weevils is due tothe absence of stimulatory substances or the in-ability to respond to such substances under certainconditions as suggested by McKibben et al (1982)

It is also of more than academic interest to de-termine the biological origin and chemical natureof the deterrent Virtually nothing is known aboutit except for its moderate insolubility in petro-leum ether The putative substance(s) could eitherbe altered or unaltered constituents of Hampeapollen or synthesized de novo by the weevil Iso-lation and identification of the active materialsfrom frass plugs themselves would be an importantstep in gaining a better understanding of weevilbiology and might also provide a powerful re-search and management tool

Acknowledgment

We thank the faculty and staff of the Colegio Superiorde Agricultura Tropical for use of facilities and for theirhelp and cooperation during the course of this and otherstudies conducted in Tabasco The following is a partiallist of persons whose remarks and criticisms we appre-ciated H R Burke A P Bhaktar O C Jenkins S JJohnston P C Krauter B F Vaissiere J K Walkerand the late W H Cross This material is based uponwork supported by USDA under Agreements No 71-59-2481-1-2-039-1 No 82 CSRS-2-1000 and No 58-519B-I-996-55290-15-4290 Approved as TA 19456 bythe Director Texas Agric Exp Stn

References Cited

Everett T H and N W Earle 1964 Boll weeviloviposition response in cotton squares and variousother substrates J Econ Entomol 57 561-566

Everett T H and J O Hay 1962 Observations onpuncturing and oviposition behavior of boll weevilIbid 55 634-642

Hedin P A H C Gueldner H D Henson and A CThompson 1974 Volatile constituents of male andfemale boll weevils and their frass J Insect Physiol20 2135-2142

Hunter H C T F Leigh C Lincoln B A Waddleand L A Bariola 1965 Evaluation of a selectedcross-section of cottons for resistance to the boll wee-vil Ark Agric Exp Stn Bull 700

Hunter W D and W D Pierce 1912 The Mexi-can boll weevil a summary of the investigations ofthis insect up to Dec 31 1911 US Dep Agric BurEntomol Bull 114

Jenkins J J W L Parrott and J W Jones 1975Boll weevil oviposition behavior multiple puncturedsquares Environ Entomol 4 861-867

McKibben G W L McGovern and W A Dickerson1982 Boll weevil (Coleoptera Curculionidae) ovi-position behavior a simulation analysis J EconEntomol 75 928-931

Mitchell H C and W H Cross 1969 Ovipositionby the boll weevil in the field Ibid 62 604-605

Sokal H H and F J Hohlf 1981 Biometry theprinciples and practice of statistics in biological re-search W H Freeman San Francisco

Received for publication 16 March 1984 accepted 5June 1984

October 1984 STANSLY AND CATE BOLL WEEVILS ON Hampea 1365

Table 2 Numbers of H nurricia flower buds accepted and not accepted for feeding by male (a) and virgin female(b) A grandi8 data arranged in contingency tables and analyzed for independence by X2

2a Accepted Notaccepted 2b Accepted Not

accepted

Uninfested

Infested

Obs 14 10Exp 12 14Obs 11 18Exp 15 17

Number of males = 3Number of trials = 2x2 = 22 P = 014

Uninfested

Infested

Obs 31 9Exp 21 32Obs 11 57Exp 35 55

Number of virgin females = 5Number of trials = 6X2= 398 P lt 00001

infested bud Hunter and Pierce (1912) noted thatonly one weevil emerged from a cotton bud re-gardless of the number of ovipositions In morerecent times it is known that some second ovipo-sit ions are successful particularly in cotton fruit(bolls) and large buds (Hunter et al 1965) It iscommon for numerous weevils to emerge from aboll which may become considerably larger thanthe bud and still be susceptible to weevils (Hunterand Pierce 1912) whereas Hampea fruit is neversuccessfully attacked In cotton fields buds arescarce early in the season while in late seasonthere are few un infested buds or bolls At suchtimes the advantages of multiple oviposition overno oviposition would select for a less discriminat-ing weevil Such selection could act by raising thethreshold at which the marking pheromone is per-ceived by or responded to by the weevil Such amechanism can explain the differences in discrim-inatory behavior observed in cotton and HampeaIf the reported contrast between modern cottonweevils and those studied at the turn of the cen-tury is real a modern evolutionary trend towardnondiscrimination may be indicated Even todaydifferences in degree of discriminatory behaviorhave been observed between weevil strains on cot-ton (J N Jenkins personal communication) Suchdifferences may explain some of the discrepanciesreported It should be noted that it is not knownwhether the weakness or lack of discriminatorybehavior in these modern cotton weevils is due tothe absence of stimulatory substances or the in-ability to respond to such substances under certainconditions as suggested by McKibben et al (1982)

It is also of more than academic interest to de-termine the biological origin and chemical natureof the deterrent Virtually nothing is known aboutit except for its moderate insolubility in petro-leum ether The putative substance(s) could eitherbe altered or unaltered constituents of Hampeapollen or synthesized de novo by the weevil Iso-lation and identification of the active materialsfrom frass plugs themselves would be an importantstep in gaining a better understanding of weevilbiology and might also provide a powerful re-search and management tool

Acknowledgment

We thank the faculty and staff of the Colegio Superiorde Agricultura Tropical for use of facilities and for theirhelp and cooperation during the course of this and otherstudies conducted in Tabasco The following is a partiallist of persons whose remarks and criticisms we appre-ciated H R Burke A P Bhaktar O C Jenkins S JJohnston P C Krauter B F Vaissiere J K Walkerand the late W H Cross This material is based uponwork supported by USDA under Agreements No 71-59-2481-1-2-039-1 No 82 CSRS-2-1000 and No 58-519B-I-996-55290-15-4290 Approved as TA 19456 bythe Director Texas Agric Exp Stn

References Cited

Everett T H and N W Earle 1964 Boll weeviloviposition response in cotton squares and variousother substrates J Econ Entomol 57 561-566

Everett T H and J O Hay 1962 Observations onpuncturing and oviposition behavior of boll weevilIbid 55 634-642

Hedin P A H C Gueldner H D Henson and A CThompson 1974 Volatile constituents of male andfemale boll weevils and their frass J Insect Physiol20 2135-2142

Hunter H C T F Leigh C Lincoln B A Waddleand L A Bariola 1965 Evaluation of a selectedcross-section of cottons for resistance to the boll wee-vil Ark Agric Exp Stn Bull 700

Hunter W D and W D Pierce 1912 The Mexi-can boll weevil a summary of the investigations ofthis insect up to Dec 31 1911 US Dep Agric BurEntomol Bull 114

Jenkins J J W L Parrott and J W Jones 1975Boll weevil oviposition behavior multiple puncturedsquares Environ Entomol 4 861-867

McKibben G W L McGovern and W A Dickerson1982 Boll weevil (Coleoptera Curculionidae) ovi-position behavior a simulation analysis J EconEntomol 75 928-931

Mitchell H C and W H Cross 1969 Ovipositionby the boll weevil in the field Ibid 62 604-605

Sokal H H and F J Hohlf 1981 Biometry theprinciples and practice of statistics in biological re-search W H Freeman San Francisco

Received for publication 16 March 1984 accepted 5June 1984