MARDI Res.l. 17(1)(1989): 55-68 Towards the developmentof an integrated pest managementsystem of in

[. Azhar*

Key words: Helopeltis,Integrated Pest Management (IPM), cultural control, chemical control, biological control, ants, early warning system

Abstrak Helopeltisialah perosakkoko yangpenting di Malaysia.Oleh itu pengurusannyasangatlah penting untuk mendapatkanhasil koko yang menguntungkan.Memandangkan koko ialah tanamanbaru di Malaysia penyelidikandalam pengurtsan Helopeltis masih berkurangan. Kebanyakan pendekatanpengurusan perosak ini berpandukanpengalaman di Afrika Barat, keranabanyak penyelidikan telah dijalankandi negaratersebut. Namun demikian,cara pengurusannyamasih berasaskan kimia, yang menjadipendekatan utama dalam pengurusan Helopeltis di Malaysia. Perkembangankawalan kimia dibincangkan. Perkara-perkarautama dalam pengawalan biologi dan kultur juga dibincangkan.Kecetekan kefahaman terhadap prinsip-prinsip yang merangkumikaedah kawalan tersebut menyebabkan kaedah ini jarang digunakandalam pengurusan Helopeltis. Namun demikian,peranan utama semutdalam mengurangkanserangan Helopeltis, terutamanya dalam ekosistemkoko dan kelapapatut diambil kira dalam menggubalsistem pengurusanperosak bersepadu (PPB) untuk Helopeltistidak kira samaada untuk kawasanpekebun kecil ataupunladang besar. Sebagaimatlamat utama untuk mengurangkanpenggunaan racun kimia, pengawasanpopulasi Helopeltis dan kefahamanterhadap ekologinya telah menjadipenting dalam pelaksanaan PPB. Penggunaansistem amaran awal adalahsalah satu carauntuk mencapaimatlamat PPB dan seterusnya mempertingkatkanhasil koko.

Abstract Helopeltisis a major insectpest of cocoain Malaysia.Hence its management is of prime importanceto ensureprofitable yield. Cocoais a relativelynew crop in Malaysiaand researchon the managementof Helopeltisis relatively scanty.As such,management approaches are largelybased on West African experienceswhere considerable research has been carried out. However,this managementapproach is largelybased on chemicals,which remainthe first line of defencein existingHelopeltis management strategies in Malaysia.The developmentof chemical control of Helopeltis is discussed. The salientfeatures of biologicaland culturalcontrol are alsoexamined. A lack of understandingof the principlesinvolved has resulted in minor use *Cocoa and CoconutResearch Division, MARDI Hilir Perak,P.O. Box 25,36307Sungai Sumun, Malaysia Presentaddress: MARDI, CocoaResearch Station, P.O. Box 324,91007Tawau, Malaysia Author's full name:Azhar Ismail @MalaysianAgricultural Research and DevelopmentInstitute 1989

55 Developmentof IPM systemof Helopeltis of them in Helopeltis management. Nevertheless, the significant role of ants in limiting Helopeltis attack, especially under the cocoa and coconut ecosystem, is worth considering in the development of an integrated pest management (IPM) system regardlessof the size of cocoa areas. With the ultimate aim of reducing chemical usage monitoring of the Helopeltis populations and the understanding of -their ecology have become important in the implementation of IPM. The use of an early warning system is one of the means by which the objectives of IPM may be achieved with subsequentimprovement in cocoa yield.

Introduction (Giesberger 1983) and H. bakeri Popp. Mirids are serious pests of cocoa. They and H. collaris Stal. in the are widespread throughout the cocoa (Entwistle 1972). growing areasof the world, and about 40 The biology of H. theobromae has speciesin the subfamily (the been describedby Miller (1941),Tan tribe Monaloniini and Odoniellini) are $97aa) and Azhar (1986). No detailed known to attack cocoa (Entwistle 1972). biological studies on H. clavifer are The tribe Monaloniini is the most known but a description of the life history widespread with the of P. apiformls was given by Pang (1981). exclusivelyfeeding on cocoa in the Mirids are polyphagous,and H. neotropics. The speciesof Helopeltis are theobromae,in addition to cocoa, also more important in the Old World cocoa attack sevenalternate host plants (Ahmad growing areas.Attacks by members of and Ho 1980). Smith (1978) reported 25 Odoniellini occur mainly in the Ethiopian alternate hosts of H. clavifer in Papua region, Sabah and . New Guinea but no report of alternate Sahlbergellasingularis (Hagl.) and hosts for P. apiformis has been recorded. Distantiella theobroma (Dist.) constitute Damage is primarily due to the the two most damaging speciesof this feeding of adults and nymphs on green tribe and are distributed mainly in West pods and young shoots or unhardened African cocoa areas(Entwistle 1972). stems. Platyngomiroides apiformn also In Peninsular Malaysia, Helopeltis attacks the chupon and fan branches as theobromaeMill. continuesto be the well (Conway 1971;Pang 1981). Feeding greatest concern of cocoa growers. It was by thesemirids resultsin the development first reported attacking cocoa at the of circular lesionson pods and lenticular Federal Agriculture Experiment Station, lesionson unhardened stems. Attacks on Serdangin 1939 (Miller 1941). On the young cherellesmay lead to cherelle wilt other hand, H. clavifer (Walk.) and and complete loss of the crop may result if Platyngomiriodes apiformls Ghauri are the infestation is severe. Infestation by the two important mirid speciesin Sabah these mirids may also lead to subsequent with the former first discovered in 1957 attack by fungal pathogens. Correct crop and the latter in 1962(Conway 1971;Pang loss assessmentattributable to Helopeltis 1981).However, mirid problems in Sabah has never been reported in Malaysia are currently of secondary importance perhaps because of the confounding with the discovery of the cocoa pod borer, nature of contributory factors. However, Conopomorpha cramerella (Snell.), in a 257o decrease in yield due to mirid late 1980. Other mirid speciespresent in infestations was reported in this region are H. antonii Sign. and FI. (Stapley and Hammond 1959). theivorqWaterh. in

56 I. Azhar

The recognition of the effect of mirid emulsion, nicotine sulfate, and tar and infestation on cocoa in Malaysia has led to petroleum distillate were tried with partial the institution of various control measures success(Entwistle 1966). The subsequent against this pest soon after they were use of DDT. after World War II, by discovered attacking cocoa (Miller l94l; painting branch unions or jorquettes and Conway 1971).Since mirids are more other mirids restingplaces provided better important in PeninsularMalaysia than in control (Collingwood 1971).The use of Sabah, this review will emphasizethe HCH evolved after the chemical was managementof H. theobromae,although proved to be more successfulthan DDT in reference to other mirid specieswill also controlling cocoa mirids (Stapley and be made where appropriate. Hammond 1959).The effectivenessof Various management approaches to HCH was attributed to its fumigative Helopeltis have been described and effect (Raw 1959). Since then, the use of reviewed by a number of authors (see this insecticidehas become widespread e.g., Miller 1941;Conway 1971;'Tan throughout the West African cocoa l9'l4b; Mainstone 1978; Azhar and growing areasas well as other parts of the Jamaludin 1981;Azhar 1984,1985, 1986). world (Entwistle 1972; Omole and Ojo However, chemicalcontrol was and still is 1981). However, the widespread use of the most widely adopted approach by this chlorinated hydrocarbon insecticide many cocoagrowers. The extensiveuse of againstcocoa mirids has resulted in the chemicalsfor the managementof localized development of resistantstrains Helopeltis is due to several reasons. (Dunn 1963; Entwistle 1964). The Among them are quick action, easily phenomenon of insecticideresistance has availableand implemented and inefficient necessitatedcontinuous screeningof natural control factors (Azhar 1985). In insecticidesfor control of cocoa mirids in this review, it is therefore appropriate to West Africa (Prins 1965;Laryea 1967; lead the discussionwith chemical control Johnson eI al. 1970; Collingwood and followed by other control measures and Marchart 1971; Marchart 1971a,1971b; eventually a discussionof the Nwana et al. 1979; Omole and Ojo 1981; implementation of an Integrated Pest Owusu-Manu 1985). Management (IPM) programme. The first attempt to control Helopeltis was probably in in 1907where Chemical control the sprayingof sulfur and Macdonaghue's Cocoa planting in Malaysia which began mixture proved a failure (Entwistle 1966). in the early sixties,is relatively new when In Java, early attempts of insecticidal compared to West Africa. Research on control of cocoa mirids were made using the management of Helopelrls is relatively the natural insecticide,derris and lead scanty and many of the chemicals used arsenate,but with little success were adopted from trial results in West (Giesberger1983). DDT was later used to Africa. control Helopeltis in Java (Entwistle 1972). The earliest report on the use of Historical view insecticidesin PeninsularMalaysia to In West Africa, numerous control control H. theobromrzewas made by measureswith specialemphasis on Miller (1941). He demonstratedthat chemicals were tried against the cocoa dusting with pyrethrum-lime at weekly mirids sincewhen they were recognizedas intervals was effective in controlling the pestsin 1910(Dundgeon 1910;Nichol and pest. Lever (1949) found that three Taylor 1954). Spraying of trees with dustingsof 0.5Vo DDT applied every 10 chemicalssuch as kerosene/soap days could check the damage causedby

57 Development of IPM system of Helopeltis

H. cinchonaeMann. and H' bradYi Table 1. Some of the commonlY used Waterh. which were attackingthe shoots insecticidesin Helopeltis control of tea bushesin the Cameron Highlands. Common names Group earliestreport on the use of The Gamma-HCH (lindane) organochlorine insecticideto control cocoa mirids in Propoxer carbamate Sabahwas in 1959when HCH dust was Dioxacarb carbamate applied to the young growing points of Cypermethrin pyrethroid pyrethroid cocoatrees (Conway 1971)' Further Alphamethrin Deltamethrin pyrethroid reports on the developmentof insecticide usageagainst cocoa mirids in Sabahdid not appearuntil the seventies. economlcconstralnts. Dusting with gamma-HcH dust has Current situation been recommendedin the Past in Insecticides Spurred by the significant PeninsularMalaysia (Anon. 1977)but the resultsof mirid control in West Africa method is now obsolete.The method was using HCH, this chemicalwas and is still also recommendedfor mirid control in widely usedby many cocoagrowers in PapuaNew Guinea (Smith 1979). PeninsularMalaysia. However, because Sprayingis the most widely practised of the reported developmentof mirid method of applying insecticidesagainst resistantstrains and the suspectedtainting cocoa mirids and the operation may be effect through extensiveuse of HCH, carried out using manual knaPsack alternativesto HCH were sought. The sprayersor portable motorized mist Department of Agriculture recommended blowers. The manual knapsacksprayers the use of other chemicalsincluding are suitablewhere cocoa trees are not too dieldrin and trichlorphon for Helopeltis tall whereasmist blowers are more control (Anon. 1977).However, dieldrin suitablefor taller trees in order to reach is no longer recommendedbecause of its pods in the upper canopy (Taylor et al. high mammaliantoxicity. Other 1982).The manual knapsacksprayers are insecticidessuch as chlorpyrifos and more popular among smallholders methamidophos,though not officially becauseof the small acreageand their recommended,were also used bY a limited resourceswhile many plantations number of cocoa growers (Azhar 1985). use the motorized mist blowers as well as Screeningefforts by Jamaludin(1981) and manual knapsacksprayers. Lee et al. (1986)further added to the list Comparatively,the motorizedmist of suitable insecticidesfor Helopeltis blower hasa higherworking rate than the control and include the carbamateand manual knapsacksprayer (Taylor et al. syntheticpyrethroid groups. A list of 1982).A worker can cover only 2-3 ha/ insecticidescurrently recommendedfor day using the knapsacksprayer compared used againstHelopeltis in Peninsular to the 5-6 ha/dayusing the motorized Malaysiais given in Table l. mist blower. At the sametime, the manual knapsacksprayer is a high volume Application methods Various application method compared to the insecticidalapplication methods are motorized mist blower which is low to employed, some of which were reviewed medium volume. Hence, larger amounts by Khoo et al. 1983.Growers may use of insecticidesare applied per unit area methods ranging from dusting, spraying using the former sprayer. and foggingdepending on the severityand Fogging is commonly used when distribution pattern of the infestation,the greater coverageof insecticideis required phenologyof cocoa crops and the and labour is limited. The operation may

58 I. Azhar be carried out using portable pulsejet 1970). As the objective of a spraying foggerswhen the infestationis not programme is essentiallyto protect the widespreador tractor-drawn pulsejet developing young cherelles,such a foggerswhen larger outbreaks occur programme that coincide with this (Taylor et al. 1982).However, fogging is particular period is appropriate. In any suitablefor use with insecticideshaving a case, each spraying programme would fairly high degree of thermostability and essentiallyconsist of twc spray only on mature cocoa with a closed applicationsabout 18-27 days apart canopy. The use of insecticideshaving (Azhar 1986;Wills 1986),with the later fumigative effect is an added advantage. application aimed at destroying Localized treatment of infestation is individuals that emerged from eggsthat encouragedbecause of the heterogenous were protected during the earlier spray distribution pattern of Helopeltisin the (Entwistle 1972 Azhar 1986). fiefd (Conway 1971;Tan 1974a;Wills Proper application timing during the 1986).In this situation, knapsacksprayers day is another factor that determines the and mist blowers are commonly used. eflectivenessof the sprayingprogramme. Taylor et al. (1982)reported that portable Based on a study of the daily activity pulsejet foggersmay also be useful for pattern of H. theobromae, Azhar (1986) spot treatment, as the application cost is suggestedthat sprayingor fogging is more only one-fourth of that of knapsack desirablein the early morning or early sprayersor mist blowers. During outbreak evening as more individuals are exposed situations,the larger pulsejet foggers during these periods. However, other could provide more rapid coverageand factors such as practicality and weather better bioefficacythan knapsacksor mist conditionsmay influence appropriale -faylor blowers.When foggingequipment is not timing. et al. (1982) reported that available,spraying with chemicalshaving the effectivenessof fogging was enhanced a strong fumigative effect at every second when appfied during 1930-2234 h as the or more row of cocoa trees may be rise of fog was suppressedby temperature practised.This method has been practised inversion conditions durins these hours. in PapuaNew Guinea (Smith 1981). Cultural control Interval and timing of Some cultural practicesin cocoa appfication ITektpeltis population production have become establishedthat build-up occurs at diff'erenttimes of the they may not be considereda part of the year with the trend being dictated by the total pest management package.These availability of food, breeding sites, and include such practicesas the management weather (Tan 1974a; Azhar et al. 1983; of shadelevel, pruning and the planting of Azhar 1986).Since the population resistantvarieties. increaseswhen pod numbers are high and after the onset of the rainy season,control Management of shade level of the pest may be achievedthrough The role of shade in the management of application of insecticidesduring the Helopekis is debatable. To date, no period of increasingpod numbers and quantitative studies have been made to ascending rainfall (Tar 1974a; Azhar ascertainthe role of shade in the 1986)or during the preceding dry season managementof Helopelrrsin Malaysia (Azhar 1986).Dry seasonspraying has except for isolated observationalreports. been recommendedin and has Helopeltisis a shade-lovinginsect, that been found to delay and reduce the build- prefers areaswith well-developedclosed up of mirid populations (Youdeowei canopieswhich afford protection from the Development of IPM system of Helctpeltis sun, heavy rain, and strong wind; or Resistant clones areaswith high humidity (Miller 1941; Researcheffort on this aspectof pest Azhar and Jamaludin1981). Smith managementof Helopehis in Malavsia is (1979),in his reviewon the scanty.The reasonsfor the lack of studies interrelationshipbetween shade types and could be due to the perennialnature of pest problems in Papua New Guinea, cocoa and the heterogenousinfestation of stated that mirid population can be ltelopeltis in the field. Currently. there damagingunder a nil or short shade are no recommendedclones which are less regime but lesstroublesome under the susceptibleto Helopeltisbut researchis in taller shadetypes. In West Africa, mirid progressto indentify cloneswhich are less infestationcan be particularly damaging susceptibleto this pest through visual where there are gaps or breaks in the assessment(Azhar 1984).Elsewhere, shadetree canopy or where overhead limited amounts of researchon the shadeis sparse(Williams 1953;Entwistle susceptibilityof clonal cocoa to mirids re72). have been reported in Ghana (Posnette In Malaysia, however. no obvious 19,13),Cameroon and lvory Coast relationshipcould be establishedbetween (Decazyand Lotode 1975;Decazy and shade types and Hebpehis attack as Coulibaly 1981)and PapuaNew Guinea infestationsoccur regardlessof whether (Smithand Moles 1981). cocoa is planted under coconut or Gliricidia. Nevertheless,the incidenceof Hand collection and scorching Helopeltis attack was observed to be less This method of Helopeltiscontrol has in cocoaplanted under taller shaderegime been practisedsince the beginningof this (such as under tall coconut of more than centuryin Indonesia(Giesberger 1983). 40 yearsold) than those planted under Hand collection of both adults and shorter shaderegime (suchas under dwarf nymphsis made usinga split bamboo and semi-dwarfcoconuts or under heavy stick, with one end dipped into some shadeof Gliricidia). This phenomenon sticky fluid. Zehntner (1903)emphasized could possiblybe due to the more exposed that the operation should be carried out environment under tall shaderegime when Helopelrls populations are low tcr leadingto lessmoisture retention preventa rapid increasein the next capacity,greater exposureto sunlight and generation.Especially in the larger better aeration as compared to the low plantationsector in Malaysia,this method and heavy shadeof cocoa monoculture of Helcpeltis managementis unrealistic under dwarf and semi-dwarfor Gliricidia becauseof time and labour costs. shade. Under the latter conditions. the However, it may be usefulto greaterintensity of the green houseeffect smallholders. tends to provide a relatively warm and Becausemanual collectionis time moist atmosphere suitable for Helopeltis and labour intensive,scorching of to flourish. Conversely,Lim (1978) Helopeltis on mature pods with fire was observedthat removal of shadein a recommendedin Indonesia particularly monoculture cocoa in Sabah led to an during the high population period increaseincidence of infestationby mirids (Giesberger1983). However, this method and other pests.Accordingly, no was not popular and was discarded recommendedgeneral policy on the shortly. In Malaysia, scorchinghas never management of Helopelrrs through shade been recommended at all becauseit is manipulation could be established. impractical.

60 L Azhar

Biological control F., Euagorasplagiatus Burm., and Parasites and predators Rhinocoris marginellus Thunbg. have Little is known of the biologicalcontrol of been mass-bredand releasedin Cameron Helopeltis by parasites and predators. Up Highlands for the control of Helopeltis on to the present no parasitehas ever been tea (Corbet and Pagden1941). However, recorded attacking H. theobromae the resultof the releaseis not known. although three speciesof egg parasites, Conway (1971)also reported the presence namely Anagyrus sp. (Hymenoptera: of three of thesereduviid speciesin Sabah Mymaridae), Erythmelus helopeltis but did not describetheir significance. Gahan (Hymenoptera: Mymaridae) and Spiderswere also observedto attack both Telenomussp. (Hymenoptera: adults and nymphs of Helopeltis in the Scelionidae),have been reported to field (Azhar 1985). attack 11.cinchonae on tea in Cameron tlighlands (Ahmad and Ho 1980).In Java Ants a braconid parasite, Leiphron helopeltidis Ants are the most abundant insect in the Ferr., with its hyperparasite, Stictopisthus cocoa ecosystem.The dominant species iavensisFerr.. was recoveredfrom are generally distributed in a mosaic H. antonii nymphs (Giesberger 1983). pattern as a result of intense interspecific Attempts at releasingthe parasite competition and the influenceof light and L. helopehidlswere made in Indonesia shaderegimes (Leston 1971,1973a, but were discontinueddue to its 1973b;Room 1971; Majer 1972. 1976a, ineffectiveness.An egg parasite. 1916b Bigger 1981;Jackson 1984). Their Telenomussp. (Hymenoptera: role as a control agent of insect pestshas Scelionidae),and a nymphal parasite. been reviewedbv Clausen(1940) and Euphorus helopeltidis (Hymenoptera: Room (1973).The use of ants as a Braconidae), have been recorded biological control agent againstcocoa attacking S. singularis and D. theobromae miridswas realized in the earlypart of this in West African cocoa with an average century when cocoa growers in Indonesia parasitismof 25% and 12Voin Ghana and deliberately introduced the indigenous Nigeria respectively(Entwistle 1972). black ant, Dolichoderus bituberculatus The nymphs of Helopeltis spp. in Mayr into the cocoa fields to control Indonesia have also been reported to be Helopeltis(Giesberger 1983).Perhaps the attacked by two speciesof mermithid most extensiveinvestigations on the nematodes, Hexamermis microamphidis ecology and role of ants in limiting cocoa Steiner and Agamermis paradecaudata mirids have been carried out in West Steiner, with parasitizationranging from Africa (seee.g., Leston 1971,1973a, 5-10% (Giesberger1983). Pathogenic 1973b:Collingwood 1971; Room 1971; fungi such as Beauvaria, Hirsutella and Majer 1972, 1976a, 19'l6b; Taylor 1977, Aspergillus were occasionally recorded 1978;Jackson 1984). Control of cocoa parasitizingcocoa mirids in West Africa pestsby ants have also been developed in (Collingwood 1911). PapuaNew Guinea (Room 1973;Smith Many predators have been observed 1981).In Malaysia,the importanceof ants to have the potential of attacking in controlfing Helopeltiswas mentioned Helopeltis in the field. A speciesof by Conway (1971)and Mainstone(1978) reduviid. Cosmolestispicticeps Stal., has who referred to the successfulIndonesian frequently been encounteredbut its example. Azhar (1985) reported the ant importance is not known (Azhar 1985). speciescomposition and briefly discussed This speciestogether with the others, their roles in the cocoa ecosystem. Sycanusleucomesus Walk., Isyndus heros In West Africa, although numerous

61 Development of IPM system of Helopeltis

Table 2. Ant species(Hymenoptet:a: Formicidae) collccted from cocoa growing areas in PcninsularMalaysia (After Azhar 19t35) D o lic ho derus tho r acic us Smith Iridomyrmex sp. (Latr.\ Dolichoderus sp. (thoracicus - gp.) Paratn rhina lttngicornis Crematogaster( Acrocoelia) sP. Paratrexhina sp" Crematogaster sp. O dontoponera tronsverso (Smith ) Meranoplus sp. B othroponera trid entzta (Smith) Tetraponera rufonigra (Jerdon) Brachyponera sp. Tetraponera sp. Hypoponera sp. Anoplolepis longipes (Jerdon) Anochaetus sp. O do nto machus simi llim us SmiIh Cerupach,vs sp. Odontomachtts rxosus Smith Pachl-cond1,lasp. Odontomachus haematodesL. Leptogenys sp. Diacamma regosum (l-e Guillon) Pheidologenton sp. Technomyrme x,r/b?e.t (Smith) Poly rhachis /eavlssirnaSmith O ec o p h1:lla sm ar agdina (F ab.) Po l,-r hachi s tlblalls Smith Solenopsisgeminata F. Polyrhuchis thrinax Roger Pheitlole sp. Polv rhachi s ftector Smith Camponotus (Tanaemyrmex) rnconspicousN4ayr Pol,- rhac his blcolor Smith Camponotus gigas Po Iyr hac his riir'e.srSmith Camponotus sp. Polyrhachis sp. M yopopone castanea(Smith) Moncmorium sp. Tetramorium pacificus Mayr Trachymesopus sp.

ant speciesare known in the cocoa forage mainly on litter. Although the ecosystem,only a few are dominant and important South American ant species are noted for their important role in Wassmanniaauropunctata Roger has been limiting mirid attacks. Among the 52 reported to give good control of cocoa specieslisted by Majer (1976) speciesin mirids, the ants failed to spread outside the genera Oecophylla, Crematogaster the area of original introduction (Mire and Macromischoides are the most 1910). common. Although earlier evidenceis Successfulcontrol of Helopeltis has somewhatconflicting (Williams 1954),the been reported in Java by introducing D. presenceor absenceof the cocoa mirid biturberculalasinto cocoa plantations D. theobroma are known to be largelY provisioned with artificial nestsof dried influencedby the presenceand absenceof bananaand cocoaleaves (Giesberger 'Ihe one of these dominant ants (Majer 1972). 1983). successof this control depends Both O ecophy lla and M acr omis cho ides entirely on the availabilityof the prey on D. theobroma and S. singularis mealybug, Planococcuslilacinus (Cockl. ), and excfude Crematogasterbut neither which provide the ants with honey-dew. It have been reported to tend mealybugs has been detnonstrated that Helopeltis (Leston 1973a).These two species populations were controlled for over 40 together have been reported to confer years by maintaining large populations of adequateprotection on 657o of mature antsand mealybugs.Van der Goot (1917) and healthy trees (Leston 1973b)although reported that increasedmortality was Collingwood (1911) provided a more observedin mealybugsunattended by the conservative figure of only 15.Zc/o. black ants as opposed to those tended by Attempts at artificially spreading ants. Currently, in Indonesia biological Oecophylla in Ghana has not been control by using ants has stopPed. successful.Collingwood and Marchart In Malaysia, Azhar (1985) listed 43 (I971) attributed the failure to the relative speciesof ants collected from the cocoa stable colonies of the Oecophvllo.which ecosystem(Table 2). Only a few were

62 I. Azhar

Table 3. Potential ant speciesas biological Other ant speciesthat seem to control asents of Helooeltis provide some Helopeftis control are Species Subfarnily O eco p hy lla smaragdina F ab., Ano p Io lep is Dolic hoderus thoracicus Dolichoderinae longipes Jerd. and Crematogaster spp. Oecophylla smaragdina Formicinae Oecophylla smaragdina does not spread Anaplolepis longipes Formicinae very fast and they seem to acquire their Crematogaster spp. Myrmicinae food items on the ground. The problem with using A. longipesto control observedto have potential in controlling Helopeltis is its superior competitive effect Helopeltis (Table 3). Among these species that may drive other ant speciesout of the only Dolichoderw thoracicris Smith seemed cocoa field, and its sporadic behaviour in to be effective against Helopeltis. T\is visiting the mealybugs. Crematogaster speciesis predominantly found in the spp. mostly forage or tend mealybugs cocoa-coconutecosystem where it forms inside their gallerieson the branch and, nestsmainly in the coconut crown (Azhar therefore, any intruder on the pods may and Musa 1988).It sometimesbuild be ignored. The associationof mealybugs temporarynests in the cocoacanopy or in with other ant speciessuch as Meranoplus the dried leaf litter on the ground and often sp. may be accidentaland therefore with a trail leading to the main nest. The render little or no effect in deterrins beneficial effect of this speciesis due to its Helopeltis attack. mutualistic relationship with a few species of mealybugs(e.g. Cataenococcttshispidru Use of pheromone (Morr.), P. lilacirun and P. pacificus Cox) The presenceof an attractant pheromone which infest mainly cocoa pods and young in H. clavifer has been reported by Smith shoots.The maintenanceof large (1917) in Papua New Guinea. In populations of these ants, which are Malaysia, caged virgin femalesof mainly arboreal, in the cocoa plantations H. theobromae has been observed to may be achievedby providing additional attract adult males in the fields indicating migrating alleys from their nestsin the the presenceof sex attractant (Azhar cocoa canopiesusing materialssuch as 1982, unpublished data). Although there fallen coconut fronds. Proper shade level were no further investigationson this such as the complete canopy of cocoa aspectthere are certainly prospectsof trees should be maintained to ensure the incorporating pheromones in the build-up of mealybug colonies as they H elopeltis management programme in thrive in well-shadedfields. Although the Malaysia, especiallyas a useful mealybug populations may occasionally monitoringtool. get excessivelyhigh, the populations would normally decline as a result of The integrated pest management system dispersaland mortality inflicted by The evolution of pest management variousnatural enemies (Azhar 1985). strategies for Helopeltls in Malaysia has The use of the black ant-mealybug centred entirely on the use of chemicals. association to control Helopeltis is only However, over reliance on chemicalshas feasiblewhere cocoais interplanted under been shown to be environmentally and coconut. It may not be feasiblein economicallyunsound (Conway 1969; monoculturesituations because black ants Azhar 1985). Unexpected problems such do not thrive well probably due to the lack as outbreaks of secondarypests may of nestingplaces. In addition, competition appear following unjustified chemical pressuresimpose by other ant speciesmay applications.Although insecticide affect the black ant in this ecosvstem. resistant strains of H. theobromae have

63 Development of IPM system of Helopeltis not been detected in Malaysia (Dzolkhifli Implementation of the IPM programme et al. 1986)its eventual developmentmay The IPM system can be implemented by 'pesticide lead to a situation of treadmill' both estatesand smallholdings.However, (van den Bosch 1975).There is no the smallholdersmay not have as denying that chemicals have an important elaborate a systemas the estatessince role in the management of Helopeltis but their resourcesare rather limited. The key they have to be harmoniously to the development of this approach is incorporated into an integrated pest monitoring with the ultimate objective of management (IPM) system to ensure reducing the amount of chemical usage. maximum economic benefits with minimal Blanket spraying is common among costs incurred. IPM is a dynamic system smallholdersregardless of pest status. and may be adjusted according This is becausesmallholders are normally to variations in cocoa cultural practices or ignorant or have little information about to incorporate any new information the pest. At the same time the small area gathered peiiodically from field research. of the holding may be completely infested In Malaysia, cocoa is either planted as a necessitatingblanket spraying. However, smallholder'scrop of minimal acreageor the managementsystem may be improved as a plantation crop with areas of more if the grower carries out regular than 100 ha. In addition, cocoa may be monitoring to ensure that chemical planted either as an intercrop under application is applied at the correct time coconut or as a monoculture under to achieve increased effectiveness and Gliricidia shade. Each of these planting reduce the frequency of chemical systems certainly influences the application. Weather pattern and management strategies of Helopeltis as cropping phenology should also be the environmental variables within each consideredin deciding chemical system will differentially affect the application. It has been found in various ecology and behaviour of Helopeltis. studies (Tan 1974a;Azhar 1986;Wills However, the following underlying 1986)that the best time to apply chemicals elements of IPM as listed by Flint and van is when there is maximum build-up of den Bosch (1982) do not vary: Helopeltis which coincide with ascending o Decision maker, in this caseeither rainfall and increasingpod numbers. Spot the owner or the manager treatment of only the infested trees could O Knowledge or information about the spare some of the beneficial pest available within the holding. Although o Sampling or monitoring this approach is slightly laborious it may o Decision making criteria be compensatedby the small size of the O Management tactics holding. The availability of black ants, especiallyunder the cocoa-coconut Detailed elaboration of these elements ecosystem, further necessitatesspot were given by Flint and van den Bosch treatment. Nevertheless,theeffectiveness (19U) and Smith (1983). Examples of of this managementsystem may not be successfulimplementation of IPM against lastingdue to the immigration of breeding Helopeltis are provided by the system of adults from the surrounding, unmanaged mirid control practised in the Lubolela holdings. As such, the institution of a plantation in the Congo Republic regional approach to Helopeltis (Entwistle 1972), Pabatu Estate in North management is often necessaryto achieve , Indonesia (Youdeowei and persistent effectiveness which is Toxopeus 1983) and Dunlop Sagil Estate, economically and environmentally Johore (Wills 1986). justifiable. This requires cooperative

64 I. Azhar

effortsfrom all growerswithin the region. of the naturally occurringbiological In the estatesector, pest controlagents, such as ants,in this managementsystem is often better system. implementedthan in the smallholder The organizationof this pest sector.Here, pest and diseasesituations managementsystem has been outlined by are routinely monitored and infestations (Wignyosoemartoand Soebiyakto1980; are promptly treated.The treatment Youdeoweiand Anwar (1983)and Wills methods,whether fogging, spot treatment (1986).Generally, the total cocoaarea is or blanketspraying depend on the divided into sectionsor blocksof severityof infestation.Since monitoring is proportionatesize which are then mainly on the basisof detectingthe subdividedinto plots of 5 ha (Figure1). presenceor absenceof damage,the Monitoring is carried out within subplots detectionof damageis occasionallylate consistingof an appropriatenumber of leadingto heavychemical application to cocoatrees to allow for satisfactory suppressthe high populationcausing the identificationof aggregationsas usedby damage.As such,the systemis also Wills (1986).A subplotconsists of a group inefficientin termsof cost-effectiveness. of four treesdistributed evenly at 5.5 This leadsto the developmentof a groupsper hectare.Wills (1986)carried managementusing the Early Warning out monitoring at 18 daysinterval as the System(EWS) or Early ResponseSystem life cycle of Helopeltis from egg to adult is (ERS). about 21 days.He alsosecured the actual number of Helopeltisat a hand height Early warning system usinga scoringsystem of 1 to 5; a scoreof This approachwas initiated and practised 1 indicatesone insect,2 indicatestwo, and in a North Sumatrancocoa estate so on but a scoreof 5 indicatesthat five or (Wignyosoemartoand Soebiyakto1980; more than five insectshave been Youdeoweiand Toxopeus1983). The use recorded.He advocatedimmediate of this systemin Malaysiais as recent as chemicaltreatment only to subplotsand 1986when Wills reportedthe adoptionof adjacentplots with a scoreof 3 to 5. this systemat the Dunlop SagilEstate, However, he did not take into Johore. Basically,the objectiveof this considerationthe presenceof ants (such approachis to reducethe amount of as their distribution) as a natural control chemicalsused without reducingits effectiveness.In this systemthe areasof infestation or aggregationof Helopeltis are locatedand treatedin the early stages of populationbuild-up limiting the outbreak.The developmentof this system takesinto considerationvarious factors involving the biology of Helopeltis, such as the rapid build-up of its population (Youdeoweiand Toxopeus1983), its aggregationbehaviour (Wills 1986)and the conditionof the cocoacanopy (Youdeoweiand Toxopeus1983). Azhar Figure I. Implementationprocedures of EWS. (1986)also suggested the incorporationof Demarcationof the area into main plots of 5 ha the crr:ppingcycle or pod numbersand each (A) and establishmentof subplot (B) weatherpatterns, especially rainfall, as wheresampling will be carried out on well asconsideration on the existingstatus predeterminednumber of trees

65 Development of IPM system of Helopeltis agent in the cocoa field although they are - (1985). Natural and cultural control in the selected cocoa pests. In known to provide some control of management of Integrated Pest Management in Malaysia (Lee, in the field (Azhar 1985). Helopeltis B. S..Loke, W. H.. and Heong.K. L.. ed.)p. The use of this sYstemhas been 255-6'7. Kuala LumPur: MAPPS demonstrated to reduce the amount of - (1936). Seasonaland daily activity patterns of insecticide and labour required in addition Helopeltis theobromae Mill. (: implications in field control. to a reduction in yield loss due to ) and their ln Cocoa and coconuts: Progress and outlook (Youdeowei and Toxopeus Helopeltis (Pushparajah,E. and Chew, P. S', ed.) p. 1983;Wills 1986).Wills (1986) indicated 305-16. Kuala Lumpur: Inc. Soc. Pltrs that only 57o of the total area at each Azhar, I. and Jamaludin, S. (1981). Insect pestsof inspectionround required treatment. In cocoa and their control. Proc. Sem. teknol' lntan, 1l-13 Nov' North Sumatra, Youdeowei and koko masakini I, Telok 1981 p. 171-86. Kuala LumPur: MARDI (1983) reported the use of , Toxopeus Azhar. I. and Musa, M. J. (1988). Studieson the chemicals was reduced by 50Vo through faunal and floral communities in the debris of the implementation of this system.As a the coconut crown. Proc. Nat. coconut conf., result of the reduction in insecticide usage Kuala Lumpur, 25-26 May 1982,p. 411-34. through spot treatmerrt, the ecology of the Kuala LumPur: MARDI Azhar. L. Musa, M. J. and Lee, C. H' (1983).Some is maintained and the natural ecosystem preliminary observations on the ecology of enemiesspared. In addition, the selected cocoa pest in West Malaysia. Proc. establishment of this system allows for the MAPPS Sem. adv. cocoa pl. prot. in Malaysia. monitoring of other aspectsof cocoa MAPPS Newsl. 7 (suPPl.): 15 (1986). growing such as diseasesand crop level. Bakri. A. H. and Redshaw, M. J. Pemberantasan Helopeltis secara terpadu problems are uncovered appropriate As dengan penggunaan semut dan bahan kimia remediescan be immediatelY pada tanaman coklat di Sumatra Utara. implemented. Disajikan pada Temu llmiah entomology perkebunan Indonesia, Medan. 22-24 Apr. Acknowledgements 1986 Bigger, M. (1981). Observations on the insects fauna grateful to Dr Richard Zack, I am of shaded and unshaded Amelonado cocoa. Department of Entomology, Washington BuIL Entomol. Res. 71: 101-20 State University and two anonymous Clausen, C. P. (1940). EntomophaSou.sinsecls. 680 reviewers for their helpful comments to p. New York: McGraw-Hill capsids in West improve the manuscript. This paper was Collingwood, C. A. (1971). Cocoa Africa: Report of international capsid research presented at the 2nd originally team t965-1971, 90 p. London: International International Conference on Plant Office of Cocoa and Chocolate Protection in the Tropics, Kuala Lumpur, Collingwood, C. A. and Marchart, H. (1971). March, 1986. Chemical control of capsids and other insect pests in cocoa rehabilitation. Proc. 3rd Int. cocoa res. conf., Accra, Ghana,23'29 Nov. References 1969,p.89-99. Lagos: Cocoa Prod. Alliance Ahmad, Y. and Ho, T. H. (1980).List of economic Conway, G. R. (1969). A consequence of pests, host plants, parasites and predators in insecticides - pest follow the chemical in the 153. West Malaysia (1920-19'18). Bull. No. cocoa of Malaysia. Nat. Hist. 78: 46-51 Agric., M'sia 538 p. Kuala Lumpur: Min. of - (1971). Pest of cocoa in sabah and their p. Anon. (1977). Cocoa, Agric. Leafl. No.53, 33 contol. 125 p. Kota Kinabalu: Min' of Agric' Kuala Lumpur: Min. of Agric. M'sia. Fish. Sabah Azhar, I. (1984). Approaches to selection and Corbet. C. H. and Pagden.H. T. (1941)'A review to the development of clonal cocoa resistant of some recent entomological investigation and cocoa pod borcr, Conopomorpha cramerella observations. Malay. agric. J. 29:341-'75 Proc' (Snell.), and Helopeltis spp. in Malaysi?. Decazy, B. and Coulibaly, N' (1981). Behaviour of 1984 Nat. cocoa breeding workshop. Serdang, cocoa cultivars with respect to biting-sucking Kuala Lumpur: MARDI (in Press) insects: Possibility of early selection to tolerant

66 I. Azhar

cocoa. Proc. Sth Int. cocoa res. conf., Cocoa Prod. Alliance Cartegena, Columbia, 18-23 Oct 1981, p. Lee, C. S., Woo, Y. K. and Hoo, C. H. (1986). Field 685-8. Lagos: Cocoa Prod. Alliance evaluation of cypermethrin (Ripcord) on the Decazy, B. and Lotode, R. (1975). Behaviour of control of some major cocoa pests. p.347-52. hybrid families of cocoa trees subjected to Refer Azhar (1986) Helopehis Sgn. attacks. Cafe Cacao 19:303-6 Leston, D. (1971). Ants, capsids and swollen shoot Dundgeon, C. C. (1910). Notes on two West African in Ghana: Interactions and the implications for Hemiptera injurous to cocoa. Bull. Entomol. pest control. p.205-21 . Reler Collingwood Res.I:59-62 and Marchart (1971) Dunn, J. A. (1963). Insecticide resistance in cocoa - (1973a). The ant mosaic: a fundamental capsid, Distantiella theobroma (Dist.). Nature property of cocoa farms. Proc. 4th Int. cocoa 199:1207 res. conf., Trinidad, 8-18 Jan. 1972, p. 570- Dzolkhifli, O., Khoo, K. C., Rita, M. and Ho, C. T. 81. Lagos: Cocoa Prod. Alliance (1986). Preliminary study of resistance in four (1973b). The ant mosaic - tropical tree populations of Helopeltis theobromae Miller crops and the limiting of pests and diseases. (Hemiptera: Miridae) to gamma*HCH, PANS 19: 311-41 propoxur and dioxacarb.p.317-23. Refer Lever, R. J. A. W. (19a9). The tea mosquito bugs Azhar (1986). (Helopeltis spp.) in the Cameron Highlands. Entwistle, P. F. (1964). The distribution of mirid Malay. agric. l. 32:91-109 speciesand of resistant mirids in Nigeria. Proc. Lim, D. H. K. (1978). New development in shade Conf. on mirids and other pests of cocoa, for hybrid cocoa in Sabah. Proc. Int. conf. Ibadan, Nigeria, 1964, p.9-17 cocoa coconut, Kuala Lumpur, 1978, p. - (1966). Cocoa mirids, Part 2. Their control. 122-41 Cocoa Growers' Bull.6:77-22 Mainstone, B. (1978). Crop protection in cocoa in - (1972). Pestsof cocoa.779 p. London: Malaysia. The Planter 54:643-58 Longman Majer, J. D. (1972). The ant mosaic in Ghana cocoa Flint, M. L. and van den Bosch, R. (1982). farms. Bull. Entomol. Res.62:151-60 Introduction to integrated pest management. - (1976a). The influence of ant and ant 256 p. New York: Plenum Press manipulation on the cocoa farm fauna. J. Giesberger, D. (1983). Biological control of the Appl. Ecol. 13: 167-75 Helopeltis pest in Java. ln Archives of Cocoa - (1976b). The maintenance of the ant mosaic Research VoI. 2. Cocoa Research in Indonesia in Ghana cocoa farms. L Appl. Ecol. 13: (Toxopeus, H. and Wessel,P. C., ed.). 123-44 Wageningen: American Cocoa Research Marchart, H. (1971,a).Chemical control of cocoa Institute and IOCC capsids: alternative to lindane. p. 173-85. Jackson, D. A. (1984). Ant distribution patterns in a Reler Collingwood and Marchart (1971) Cameroonian cocoa plantation: investigation - (1971b). Evaluation of insecticides for the of the ant mosaic hypothesis. Oecologia 62: control of cocoa capsids in Ghana. FAO Pl. 318-24 Prot. Bull. 19:97-109 Jamaludin, S. (1981). Screeningof insecticidesfor Miller, N. C. E. (1941). Insectsassociated with Helopeltis control. Ann. Rep. 1981. Cocoa cocoa () in Malaya. Bull. Coconut Div., MARDI (mimeo.) Entomol. Res. 328'. 1-16 Johnson,C. G., Burge, G. A. and Gibbs, D. G. Mire, P. B. D. (1970). Une fourmi utilisee au (1970). Field trials of anti-capsid insecticides Cameron dans la lutte contre ies mirides du on farmer's cocoa in Ghana, 1950-1960. cacaoyer Wasmannia auropunctaRoger. Cafe' 1. Comparing the effect of treatments by Cacao 13:209-12 assessingsubsequent damage. Ghana J. agric. Nichol, J. and Taylor, D. J. (1954). Capsids control Sci.3: 155-78 in the Belgian Congo with special reference to Khoo, K. C., Ho, C. T., Ng. K. Y. and Lim, T. K. Lukolela plantations. West African Cocoa (1983). Pesticide application technology in Research Institute, Tech. Bull. No. 2, l0 p perennial crops in Malaysia. ln Pesticide Nwana, I. E., Adesiyan, S. O. and Youdeowei, A. application technology (Lim, G. S. and (1979). Evaluation of some new pesticides for Ramasamy, S., ed.) p. 42-85. Kuala Lumpur: the control of cocoa mirids: Laboratory MAPPS assessments. Proc. 7th Int. cocoa res. conf., Laryea, A. A. (1967). Field trials with carbaryl Doula. Cameroon. 5-12 November 1979. againstcyclodience resistant mirids on cocoa in p. 453-55. Lagos: Cocoa Prod. Alliance Ghana. Proc.2nd Int. cocoa res. conf., Bahia, Omole, M. M. and Ojo, A. A. (1981). Field trials Brazil, 19-26 Nov 1967. p. 296-303. Lagos: with insecticides to control cocoa mirids

67 Development of IPM system of Helopeltis

Sahlbergella singularis Hagl. in Nigeria. p. population fluctuations. Malay. agric. Res. 3: 339-43. Refer Decazy and Coulibaly (1981) tlt--)z Owusu-Manu, E. (1985). The evaluation of - (1974b). Helopehis theovora theobromae on synthetic pyrethroids for the control of cocoa in Malaysia. II. Damage and control. Distantiella theobroma (Dist.) (Hemiptera: Malay. agric. Res. 3:204-12 Miridae) in Ghana. Proc. 9th Int. cocoa res. Taylor, B. (1977). The ant mosaic on cocoa and conf., Togo. Ghana, 12-18 February 1984. other tree crops in Western Nigeria. Ecol. Lagos: Cocoa Prod. Alliance Entomol. 2:245-55 Pang, T. C. (1981). The present status of cocoa bee - (1978). The abundance and inter-specific btg, Platyngomiriodes apiformis Ghauri, in relations of common ant species Sabah and its life cycle study. Proc. Int. Symp. (Hymenoptera: Formicidae) on cocoa farms in problem of Insect Pest Management in Western Nigeria. Bull. Entomol. Res. 68'. developing countries, Kyoto, Japan, 1980. 105-21 Trop. Agric. Res. Ser. No. 14 p.83-9O Taylor, J. G., Ho, C. T. and Lim, T. M. (1982). A Posnette. A. F. (1943). Resistance of Theobroma mechanized thermal fogging system for cacao to Sahlbergella spp. on the Gold Coast. controlling Helopeltis theobromae Mill. in Bull. Entomol. Soc. 34:159-62 cocoa. In Proc. Int. conf. on plt. prot. in the Prins, G. (1965). Contact toxicities of 22 insecticides tropics. p.695-703. Kuala Lumpur: MAPPS to the cocoa mirids Drsranriella theobroma van den Bosch, R. (1975)., Pesticidesconspiracy. (Dist.) (F{emiptera: Miridae). Bull. Entomol. 226 p. New York: Doubleday and Co. Inc Res. 56:231-5 van der Goot, P. (1971). The black cocoa ant, Raw, F. (1959). Studies on the chemical control of Dolichoderw bituberculatus Mayr. and its cocoa mirids. Distantiella theobroma (Dist.\ importance for the cocoa culture in Java. and Sahlbergella singularis Hagl. Bull. Mededlingen Proefstation Midden Java 25: 1- Entomol. Res. 50:13-23 142 Room, P. M. (1971). The relative distributions of Wignyosocmarto, S. and Soebiyakto, K. A. (1980). ant species in Ghana's cocoa farms. J. Appl. Pendalian hama tanaman cokelat dengan Ecol. 40:735-51 metoda pengamatan dini (early waming - (1973). Control by ants of pest situations in system). Proc. lst Nat. cocoa conf., Medan, tropical tree crops: a strategy for research and Indonesia. 16-18 September 1980, p. 135-44 developments. Papua New Guinea Agric. J. Williams, G. S. (1953). Field observationson the 24:98-103 cocoa mirids Sahlbergella singularis Hagl. and Smith, E. H. (1983). Integratedpest management Distantiella theobroma (Dist.) in the Gold (IPM) - specific needs of developing Coast. Part I. Mirid damage. Bull. Entomol. countries. Insect Sci. Appl. 4:113-7 Res.44:101-19 Smith, E. S. C. (1977). Presenceof a sex attractant - (1954). Field observations on the cocoa pheromone in Helopeltis clavifer (Walker) mirids Shalbergella singularis Hagl. and (: Miridae). J. Aust. Entomol. Distantiella theobroma (Dist.) in the Gold Soc. 16:773-6 Coast. Part III. Population fluctuations. Buli. - (1978). Host and distribution records of Entomol. Res.45:'723-44 H eIo p eIt is clav ife r (ll alker) (Heteroptera : Wills, G. A. (1986). Use of an early warning system Miridae) in Papua New Guinea. Papua New for the control of Helopeltis theovora GuineaAgric. J.29: 7-4 theobromaein cocoa. p.241-53. See Azhar - (1979). Interrelationships between shade ( 1986) types and cocoa pest diseaseproblems in Youdeowei, A. (1970). Timing insecticides Papua New Guinea. p.37-43. SeeNwana et application to control the cocoa capsid al. (1979). Sahlbergella singularis Hagl. in Nigeria. 1nr. - (1981). An integrated scheme for cocoa pests PestContr. 12:13-76 and diseasesin Papua New Guinea. Trop. Pest Youdeowei, A. and Anwar, F. (1983).The Manag. 27:351-9 Helopeltis census and management system. Smith, E. S. C. and Moles, D. J. (1981). Bull. Cocoa Advisory Group for Indonesia 2 Susceptibility of clonal cocoa to attack by two Youdeowei, A. and Toxopeus, H. (1983) The insectpests in Papua New Guinea. p.735-9. development and benefits of the Helopeltis See Decazy and Coulibaly (1981) censusand managementsystem on cocoa in Stapley,J. H. and Hammond, P. S. (1959). Large North Sumatra. Indonesia. Proc.2nd Nat. scale trials with insecticidesagainst capsids on cocoa conf., Medan, 1983, p. 158-66 cocoa in Ghana. Emp. l. F.xp. Agric. 27: Zehntner, L. (1903). T'ne Helopeltis pest in the 343-53 cocoa culture and its control. Bulletin Tan, G. S. (197aa). Helopeltis theovora theobromae Proefstation v. Cacao. Salatiga 7: 7-22 on cocoa in Malaysia. I. Biology and

Accepted for publication on ll May 1988

68