INSECT FAUNA SURVEYS ON RAMBUTAN, DURIAN AND MANGOSTEEN IN NORTH QUEENSLAND

David Astridge

A description about Darwin and the Northern Territory’s natural landscapes, ecosystems and wildlife. I also discuss the dramatic seasons of the monsoonal north and the storms birthed within this steamy tropical environment. This is a section of the main story “Monsoon Thunder in the Land of Fire and Flood.’

A revised set of propositions about ecology in arid Australia is presented, based on research literature since publication of Stafford Smith and Morton (1990). Fourteen propositions distil our argument that most features of the Australian deserts are explicable in terms of two dominant physical and climatic elements: rainfall variability, leading to extended droughts and occasional flooding rains; and widespread nutrient poverty. Different landscapes within the arid zone show these features to varying degrees, and so it is important to think about different places separately when considering our propositions. Plant life-histories strongly reflect temporal patterns of soil moisture; because Australian deserts receive more variable rainfall than most others, there is a distinctive spectrum of life-histories. Low levels of phosphorus (together with abundant soil moisture on irregular occasions) favour plants producing a relative excess of carbohydrate (C). In turn, C-rich plant products sometimes lead to fire-prone ecosystems, assemblages dominated by consumers of sap and other C-based products, and abundant detritivores (particularly termites). Fluctuations in production due to variable rainfall provide openings for consumers with opportunistic life-histories, including inhabitants of extensive but ephemeral rivers and lakes. Most consumer species exhibit some dietary flexibility or utilise more dependable resources; these strategies give rise to greater stability in species dynamics and composition of assemblages than might first be imagined under the variable rainfall regime. Aboriginal people have had long-standing ecological influence as they accessed resources. For each proposition we suggest the extent to which it is ‘different’, ‘accentuated’ or ‘universal’ in comparison with other deserts of the world, recognising that this categorisation is in need of critical testing. Further tests of each proposition are also suggested to fill the many gaps that still exist in our knowledge of the structure and functioning of Australia’s deserts.

TECHNICAL BULLETIN NO. 288 PROCEEDINGS OF THE SIXTH WORKSHOP FOR TROPICAL AGRICULTURAL ENTOMOLOGISTS DARWIN - MAY 1998 Agdex 602 $8.80 (GST included) ISBN 0 7245 3077 0 Proceedings of the Sixth Workshop on Tropical Agricultural Entomology Darwin 11 – 15 May 1998 Published May 2001 Table of Contents INTRODUCTION ..........................................................................................................................................3 REVIEW OF HORTICULTURAL AND AGRICULTURAL DEVELOPMENTS IN THE NT FROM 1991-1998 - ESC SMITH ................................................................................................................................................5 DEVELOPMENTS IN HORTICULTURE AND ENTOMOLOGY IN NORTH QUEENSLAND SINCE 1991 HAC FAY .................................................................................................................................................... 11 BIOLOGICAL CONTROL OF WEEDS IN THE NORTHERN TERRITORY - ANDREA WILSON ............. 15 REVIEW OF THE CURRENT COTTON RESEARCH PROGRAM IN THE ORD - AJ ANNELLS AND GR STRICKLAND............................................................................................................................................. 21 ERADICATION OF THE ORCHID WEEVIL ORCHIDOPHILUS ATERRIMUS (WATERHOUSE) (COLEOPTERA: CURCULIONIDAE) FROM THE NT - ESC SMITH AND MJ NEAL ............................... 27 A REVIEW OF SEXAVA RESEARCH AND CONTROL METHODS IN PAPUA NEW GUINEA - GR YOUNG....................................................................................................................................................... 31 INTEGRATION OF CHEMICAL, CULTURAL AND BIOLOGICAL CONTROLS IN NORTH QUEENSLAND SUGARCANE FOR GREYBACK CANEGRUB, DERMOLEPIDA ALBOHIRTUM (WATERHOUSE) (COLEOPTERA: SCARABAEIDAE) - L ROBERTSON ............................................... 45 BAITS FOR FRUITPIERCING MOTHS - THE STATE OF PLAY - HAC FAY AND KH HALFPAPP ........ 51 INTRODUCTION OF TETRASTICHUS BRONTISPAE FOR CONTROL OF BRONTISPA LONGISSIMA IN AUSTRALIA - K HALFPAPP................................................................................................................. 59 BIOLOGICAL CONTROL OF PALM LEAF BEETLE, BRONTISPA LONGISSIMA (GESTRO) (COLEOPTERA: CHRYSOMELIDAE) WITH THE WASP PARASITOID, TETRASTICHUS BRONTISPAE (FERRIERE) (HYMENOPTERA: EULOPHIDAE) IN DARWIN - D CHIN AND H BROWN........................ 61 EGG PARASITOIDS OF FRUITSPOTTING BUGS (AMBLYPELTA SPP.): POTENTIAL AND LIMITATIONS OF MANIPULATIVE RELEASES - HAC FAY, SG DE FAVERI AND RK HUWER ........... 67 INSECT FAUNA SURVEYS ON RAMBUTAN, DURIAN AND MANGOSTEEN IN NORTH QUEENSLAND - D ASTRIDGE............................................................................................................................................ 75 POTENTIAL OF USING COLONIES OF THE GREEN ANT, OECOPHYLLA SMARAGDINA (F.), TO CONTROL CASHEW INSECT PESTS - R PENG, K CHRISTIAN AND K GIBB ...................................... 81 THE IPM OF SNAKE BEAN, VIGNA UNGUICULATA SSP. SESQUIPEDALIS, IN THE TOP END OF THE NORTHERN TERRITORY - GR YOUNG AND L ZHANG.................................................................. 95 IPM OF MELON THRIPS, THRIPS PALMI KARNY (THYSANOPTERA: THRIPIDAE), ON EGGPLANT IN THE TOP END OF THE NORTHERN TERRITORY - GR YOUNG AND L ZHANG................................. 101 FACTORS INFLUENCING THE SPATIAL DISTRIBUTION OF THE TEA MOSQUITO BUG, HELOPELTIS PERNICIALIS, IN CASHEW PLANTATIONSR PENG, K CHRISTIAN AND K GIBB ...... 113 THE PUZZLE OF INSECT DAMAGE IN CONVENTIONALLY CULTIVATED AND NO TILLAGE PLOTS AT THE DOUGLAS DALY RESEARCH FARM - GR YOUNG AND ESC SMITH ................................... 121 DETECTION AND ERADICATION OF THE EXOTIC FRUIT FLY BACTROCERA PHILIPPINENSIS DREW AND HANCOCK (DIPTERA: TEPHRITIDAE) IN THE NORTHERN TERRITORY - ESC SMITH 127 DETERMINATION OF HEAT TOLERANCE IN IMMATURE STAGES OF BACTROCERA AQUILONIS (MAY) (DIPTERA: TEPHRITIDAE) FRUIT FLY - H WALLACE............................................................... 133 SIGASTUS WEEVIL - AN EMERGING PEST OF MACADAMIAS IN NORTH QUEENSLAND - HAC FAY, SG DE FAVERI, RI STOREY, AND J WATSON...................................................................................... 137 NEW AND POTENTIAL ARTHROPOD PESTS RECORDED IN THE NORTHERN TERRITORY FROM 1991 - 1997 - ESC SMITH AND HH BROWN .......................................................................................... 141 BIOLOGICAL CONTROL OF THE GIANT SENSITIVE PLANT WITH HETEROPSYLLA SPINULOSA (HOM. : PSYLLIDAE) IN PAPUA NEW GUINEA - LS KUNIATA AND KT KOROWI ............................. 145 INTRODUCTION The Sixth workshop on Tropical Agricultural Entomology was held in Darwin from 11 to 15 May 1998. This publication contains contributed papers to the workshop by participants from Queensland, the Northern Territory, Western Australia and Papua New Guinea. The first of these workshops was held at Mareeba, north Queensland in 1979. The workshops are an important forum for the exchange of ideas amongst tropical agricultural entomologists. Prior to the workshop, it was the consensus amongst the potential delegates that papers from the workshop be citable publications. The responsibility for refereeing of the contributed papers has rested entirely with the authors. The subject matter covered at the workshop included entomology in relation to agricultural developments in north Queensland and the Northern Territory, fruit flies, cotton entomology, pests of horticultural crops, biological control of weeds and pests of ornamentals. It is to be hoped that agricultural entomologists retain their enthusiasm and finance to carry on the workshops. The publication of the workshop proceedings owes much to the efforts and advice of Deanna Chin, Haidee Brown, Heather Wallace and Lanni Zhang, Entomology Branch NT DPIF. Graham Young Editor and convenor 3 REVIEW OF HORTICULTURAL AND AGRICULTURAL DEVELOPMENTS IN THE NT FROM 1991-1998 ESC Smith Entomology Branch Department of Primary Industry and Fisheries GPO Box 990 Darwin NT 0801 Abstract In contrast to the large production in other States of Australia, the value of the rural industries in the Northern Territory (NT) is rather insignificant. In 1996, the total value of the rural industries totalled about $306m of which the agricultural and horticultural components contributed only 16.4%. This presentation reviews the relative values of the respective components of the rural industries in the NT; summarises the main developments in NT horticulture and agriculture over the past seven years since the Townsville workshop of Applied Tropical Entomologists; examines in some detail the development of our most important horticultural crop - mangoes; and will attempt to predict developments over the next few years. Overall, the rural industries in the NT are currently entering an interesting period with good growth potential in several sectors and a “shaking down” in other areas. Introduction The total value of four major rural industries in the NT (pastoral, fisheries, broad acre field crops and horticulture) during 1996 was about $306m, of which horticulture accounted for <16% and broad acre field crops about 0.8%. However, over the past six years, the value of the horticulture industry has increased by around 125% while the pastoral industry increased by about 50% and fisheries enjoyed a 25% increase in value (Table 1). Table 1. Values of Northern Territory rural industries ($m) 1991-1996 PASTORAL FISHERIES FIELD CROPS HORTICULTURE 1991 105 87 1.6 20.9 1992 133 90 4.0 28.9 1993 111 106 4.0 31.1 1994 170 108 3.3 41.0 1995 152 111 3.3 44.8 1996 149 107 2.4 48.0 The major areas of commercial horticultural and broad acre crop production in the NT were described in the previous workshop. They consist of three main zones: tropical coastal (centred on Darwin and extending in a semi-circle about 100 km in diameter), tropical inland (centred on Katherine and extending about 50 km around it) and arid (within 50 km of Ti tree, in Central Australia). Small areas of broad acre field crops are also grown in the Douglas Daly basin between the Darwin and Katherine regions. 5 Values of the Various Components of the NT Agricultural and Horticultural Industries The values of the major components of the NT agricultural and horticultural industries over the period 1991-1997 and estimated values for horticulture production are presented in Table 2. The data indicates that the value of grain and seed production is quite low and has decreased over the past few years. In the NT context, hay production is significant and is expected to be considerably higher in 1997 due to the very large increase in demand by the live cattle export trade. Within horticulture, the fruit sector has been the most impressive growth area over the past six years. It increased 3.2 times in value, while the vegetable sector has been quite steady (although statistics may be less accurate than for other sectors) and the value of nursery and cut flower production has fluctuated, again perhaps due to the difficulty of gathering reliable statistics. Table 2. Values of the main components in the Northern Territory agricultural and horticultural industries ($m) 1991-1997 1991 BROAD ACRE CROPS Grain Seeds Hay Total 1.3 0.2 N/A 1.6 Fruit 14.3 HORTICULTURE Veges Nursery 3.2 3.4 Total 20.9 1992 1.4 0.2 2.3 4.0 21.6 4.0 3.4 28.9 1993 1.0 0.2 2.8 4.0 24.0 3.7 3.4 31.1 1994 1.0 0.4 1.9 3.3 29.0 3.7 8.3 41.0 1995 0.8 0.3 2.2 3.3 34.1 4.0 6.6 44.8 1996 0.5 0.1 1.8 2.4 36.6 4.9 6.6 48.0 1997 N/A N/A N/A N/A 46.4 4.6 6.6 57.6 The value of NT horticultural production by sector and growing region over the past three years is presented in Table 3. The figures indicate that the increase has been mainly in the Darwin region. There is likely to be large increases in fruit production from Katherine in the next few years as large areas of newly planted asparagus, citrus and mangoes come to bearing. Table 4 illustrates the relative values of the most widely grown crops in the fruit and vegetable sectors. The estimated values for mango and banana production in 1997 again reflect the increased plantings in the Darwin region while the rockmelon production is mainly based in Katherine. Similarly, the healthy increase in bitter melon production has been achieved by a large increase in plantings in the Darwin area by new Asian growers. The destination of produce is little changed over the past years (Table 5). The excess produce is exported either interstate or overseas due to the small and scattered population centres within the NT. 6 Table 3. Value of NT horticultural production ($m) (by type and region) 1995 1996 Est. 1997 Fruit Vegetables Nursery 34.1 4.0 6.6 36.6 4.9 6.6 46.4 4.6 6.6 TOTAL By region (Fruit and Vegetables) 44.8 48.0 57.6 Darwin Katherine Alice Springs 21.4 9.6 7.1 25.6 8.6 7.2 By Type Table 4. Values of the major Northern Territory horticultural commodities 1995 1996 Est. 1997 FRUIT ($m) Mango Table Grapes Banana Rockmelon Citrus 19.8 6.4 3.0 2.5 0.7 19.4 6.1 4.4 3.4 1.3 26.0 8.1 7.2 VEGETABLES ($’000) Pumpkin Lettuce Beans Bitter Melon 485 419 403 389 549 539 474 892 Table 5. Market destination of Northern Territory fruit and vegetables 1989 1995 1996 18.3 38.2 41.6 Interstate 75.5 80.0 79.6 Overseas 5.4 7.4 7.6 NT 19.1 12.6 12.8 Total Value ($m) Percentage marketed Growth of the Mango Industry Mango is the most successful horticultural or agricultural crop grown in the NT. Prices are usually high compared with Queensland competitors because the NT harvest normally begins early and in most seasons finishes as the Queensland crop matures. The effect of high prices for early fruit explains the fact that although the NT produces about 15% of the Australian mango crop, it receives some 26% of the total Australian market value. 7 Table 6. Northern Territory mango production (tonnes) 1984 84 1987 484 1990 1,784 1993 3,731 1996 5,357 1997 8,100 The trend of annual increases in production within the NT (Table 6) is very similar to that exhibited in other mango growing areas of Australia. The table shows a steady increase during the successive three-year intervals and a massive increase in the last year when conditions were very good and many trees started bearing. The NT production is certain to further increase rapidly over the next few years since a tree survey conducted in 1997 estimated that around 62% of all trees planted in the NT were under six years of age and therefore had not reached full production. Like other horticultural commodities, a major portion of NT mangoes are marketed interstate where they attract higher prices because they are early in the market. Some 25% of NT fruit is sold to Brisbane agents, although how much of this fruit is then resold to the Sydney or Melbourne markets is unknown. Short Term Future Prospects To project the probable trends in agriculture and horticulture in the NT over the next five to ten years I will discuss the broad-acre crops situation and three horticultural sectors: fruit, vegetables and nursery/cut flowers. Short Term Future Prospects - Fruit Since many of the fruit trees planted are yet to achieve full production, we can expect increased production without additional plantings. Although pest and disease problems will undoubtedly affect the profitability of an enterprise, they will probably not be limiting for fruit production in the NT except perhaps for fruit piercing moths. However, citrus canker and other exotic pest and disease incursions will always pose a threat to horticulture, particularly in the Top End. There is an increasing interest in the adoption of at least some Intergrated Pest Management (IPM) principles for orchards and considerable local information is now available to growers who wish to follow these recommendations. The likely significant increases in horticultural production or plantings will be in the following crops and growing regions: • Mangoes: Both in the Darwin and Katherine regions. • Citrus: Katherine and to a lesser extent Darwin. There will possibly be some increase in the arid zone area if proposed plantings on aboriginal land occur. Market “windows” for lemons, limes and red grapefruit appear to offer the best prospects. • Bananas: Darwin area. Increased plantings will be limited by available suitable land. Currently there are significant areas with young plants and intended areas for future planting. 8 • Table grapes: Significant areas are still being planted in the arid zone and there is interest in Aboriginal communities. • Other tropical fruits: Darwin area. There is considerable interest in, and likely increased plantings (albeit small) of a range of species, such as mangosteen, carambola, jackfruit and durian in “mixed tree cropping” smallholder plots. The interest in large-scale cashew plantations has declined over the past few years. However, this nut crop, in addition to dates and figs in the arid zone may receive significant horticultural investments in the future. Similarly, although there have been few new rambutan plantings over the past five years, problems with production, pest and marketing are gradually being overcome. Renewed interest should be generated as the local tourist industry expands and better marketing arrangements for small quantities of specialty lines are developed in southern markets. The recent development and promotion of a management system for Mastotermes, a significant production pest in orchards, has been well received and should now be utilised by most commercial growers. Short Term Future Prospects - Vegetable Production Compared with the situation for fruit production, pest and disease problems will probably have major implications for vegetable crop production. Recently the incidence of disease has affected the viability of asparagus, bean and basil crops. Melon thrips, two-spotted mites and nematodes are major pests of many tropical vegetables and IPM practices are being developed and promoted to growers. Although few vegetable growers have, as yet, accepted IPM as a management system, the potential benefits of using high potassium soaps and the introduction of predatory mites have been demonstrated on various properties. The following crops are likely to continue developing steadily over the next few years: • Tropical/Asian vegetables - currently mainly grown around Darwin, but there is good scope for production in the Katherine region. In the Katherine area, one grower has invested heavily in asparagus production under centre pivot irrigation. Most crops have a good short-term future but there are doubts about the longer-term sustainability of land, disease control and the commitment of current growers. The marketing and pest problems appear to be becoming less important. There are good market prospects for most vegetable crops, especially asparagus, beans, bitter melons, capsicum, eggplant, various herbs, lettuce, loofah, okra, pumpkin, tomatoes, zucchini and some other specialty vegetables, such as Lebanese cucumber. • Traditional vegetables - quality produce (e.g. asparagus, beans, brassicas, capsicum, cucumbers, eggplant, lettuce, pumpkin, various melons, tomatoes and squashes) have long been grown in the Ti Tree area of the arid zone. Prospects for increased production there and on Aboriginal land for sale into market “windows” appear promising. One enterprise producing hydroponic “fancy” lettuces and other vegetables has recently discovered a very profitable niche market both locally and in the South East Asian markets. In the northern regions, growers already directing their production to markets overseas should continue to expand. The demand for tropical vegetables and good quality traditional vegetables will continue to grow. The specialty lines such as bamboo shoots, gourds, herbs, “vine-ripened” or sun-dried tomatoes and hydroponic vegetables should attract good demand and prices. As the marketing arrangements are further improved and the difficulties of vegetable production in southern areas increase due to water and land availability and environmental pressures, the prospects for vegetables grown in the NT appear very good. Short-term future prospects - Nursery and flower production The nursery industry has undergone steady growth since climatic conditions in the Top End are very suitable for most lines. Emphasis on flower production has been in the selection of suitable 9 species and marketing (e.g. gerberas, heliconias, orchids, torch and tulip gingers). Some overseas markets for native cut flowers (kangaroo paw, geraldton wax) from Central Australia have been developed but are generally in their infancy. There is scope to increase production in most lines if the quality and marketing arrangements are satisfactory. The nursery sector has already accepted the concept of IPM, particularly the use of “softer” chemicals and introductions of beneficial organisms following research and surveys conducted several years ago by DPIF. Short Term Future Prospects - Broad acre farming Although broad acre farming has declined in the Top End in recent years (as shown in the figures presented earlier in Table 2), the tremendous growth in the live cattle market especially during the period 1994-1997, promoted hay and pasture seeds as valuable commodities within a sustainable tropical ley farming system. Grain production has declined to almost negligible levels. However, further investment in centre pivot irrigation systems will boost grain legume production, particularly peanuts and soybean. There are good prospects for high value sesame and seed crops. Conclusion We are likely to be in an exciting period of horticulture and broad acre farming over the next few years as increased research in tropical horticulture and the Ord River Stage 3 development proceeds. The establishment of the proposed CRC for Tropical Horticulture involving the CSIRO, universities and state departments of primary industries would ensure that applied research and technological developments are conducted in tropical regions. There is likely to be increased investment interest in bananas, citrus, exotic fruits, figs, specialty tropical vegetables and in higher value broad acre crops, such as peanuts, cotton, sugar cane and grain legumes. The Keep River area within the Ord River Stage 3 development could provide suitable land for crops of cereals, vegetables, grain legumes, pastures grown for seed production and some fruits. As the live cattle trade improves, the demand for improved pastures and hay will again be high. However, production will be constrained by physical (e.g. climate, soil type, nutrition, reliable water quantity and quality, new technology), human (e.g. land tenure, availability of finance, management limitations, market opportunities), biological (e.g. suitable varieties, insect pests, diseases, vertebrate pests) and other factors. 10 DEVELOPMENTS IN HORTICULTURE AND ENTOMOLOGY IN NORTH QUEENSLAND SINCE 1991 H.A.C. Fay Queensland Horticulture Institute Queensland Department of Primary Industries PO Box 1054 Mareeba Queensland 4880 Abstract The gross value of horticultural production in Queensland reached the $1 billion mark in 1993/94, with the northern part of the state contributing around 40%. While horticulture has continued to grow and diversify in the region, it was significantly disrupted by the detection of the Asian Papaya fruit fly, Bactrocera papayae Drew and Hancock, near Cairns in October 1995. New requirements were imposed on growers for treating or evaluating fruit prior to shipment from the Quarantine Area (north of 19oS). A campaign to eradicate the fly was initiated, and supported by a national cost-sharing arrangement under SCARM. The campaign was due for successful completion in August 1998, after expenditure approaching $35 million. Despite the impositions growers faced, there have been substantial increases in production in most commodities in recent years; particularly in the mango, papaw, bean and banana industries. Significant crop diversification on the Atherton Tableland has seen the number of commodities worth $1 million p.a. reach 13. However, horticulture in north Queensland has had to face other developing entomological problems. These include silverleaf whitefly, Bemisia tabaci (Gennadius) type B, in vegetable and melon crops, spider mites in papaws, sugarcane weevil borer, Rhabdoscelus obscurus (Boisduval), in palms, a Sigastus weevil in macadamias, and a Cryptophlebia sp. in avocados. Among several incursions of exotic pests into Torres Strait and onto the mainland, two of the most significant have been the arrival of spiralling whitefly in Cairns in 1998, and detection of mango leafhopper near Weipa in 1997. Key words: Tropical fruit, vegetable and melons, Bactrocera papayae, Bemisia tabaci Introduction Horticultural production in Queensland has continued its rapid growth, with the Gross Value of Production (GVP) exceeding the $1 billion mark in 1993/94 (Source: ABS). North Queensland contributes around 40% of Queensland’s horticulture GVP, and crop diversification has remained a feature of the region through the 90s as circumstances in other industries (e.g. tobacco, rice) have changed. The most significant event for horticulture in north Queensland since 1991 was the detection of the Asian Papaya fruit fly, B. papayae Drew and Hancock, near Cairns in October 1995. Following a declaration of a 75,000 km2 Quarantine Area, produce from the region north of 19oS had restrictions placed on it, requiring certain conditions be met before it could be moved. In crops such as mangoes and papaws this meant dipping fruit in dimethoate or fenthion, in bananas only hard green fruit could be shipped, or non-host status had to be determined. The newly established mango trade to Japan, based on vapour-heat treated fruit, was suspended pending data on the treatment relevant to B. papayae. These data were collected and the trade re-established within a year. Nevertheless, losses to the region due to the incursion were estimated to exceed $100 million in the year following detection. An eradication campaign, based on the Male Annihilation Method, was launched in November 1995 (Fay et al. 1997). It was supported by national funding through the Standing Committee for Agriculture and Resource Management. At a cost, which approached $35 million, the campaign was due for completion in August 1998, after no fly detections for over 12 months. 11 Commodity Production Trends Figure 1 indicates the percentage increase/decrease in production in 10 major crops in north Queensland between 1991 and 1996. In bananas and papaws production increases have largely been a result of increased plantings, while in mangoes and macadamias they can be attributed to tree maturation. Pumpkins are one of the few commodities to show a decline in production, despite the expansion in varietal offerings. Figure 1. Changes in production and yield in ten major crops in north Queensland between 1991 and 1996 Other production trends include: Avocados: Few changes are occurring. A niche market has been created for early season Shepherds from the Atherton Tableland. Very little product is exported. Bananas: Plantings in southeast Queensland are declining, but those in north Queensland have increased over the past two years. While most expansion has occurred on the wet coast, there have also been developments on the Atherton Tableland. Capsicums: The Bowen area produces about 80% of the Queensland crop. Moderate increases in production are expected, with more processing. Macadamias: Expansion in the industry in Queensland has been running at about 12% p.a. Exports account for 64% of the crop. Plantings in north Queensland are static. Mangoes: Over the past five years 50% of plantings throughout the State have been of the later varieties Keitt, R2E2, Palmer and Nan doc mai. Exports, mainly from north Queensland, have risen to 10% of the crop. Over the next five years production in central and southern Queensland is expected to increase from 9% to 40% of the state level. 12 Melons: Any future expansion in the industry will be exports from north Queensland, but the economic situation in Asia will drive this. Papaws: About 90% of production is now derived from north Queensland, after recent rapid declines in the industry in other parts of the State. There has been some diversification amongst sugarcane and banana producers into pawpaw production. Tomatoes: Bowen remains the main production area in the State. No production increases are expected and grower numbers should decline. Horticulture has significantly diversified and expanded in the Mareeba-Dimbulah Irrigation Area on the Atherton Tableland, where historically tobacco has been the dominant crop. Horticultural crops worth more than $1 million p.a. have increased from four in 1990/91 to 13 in 1996/97 (including macadamias, tea tree, longans, lychees, citrus, sweet potato and flowers) (Anon 1998). The value of the MDIA’s lychee industry alone has grown to over $10 million p.a. However, sugarcane production is also expanding rapidly on the Atherton Tableland, occupying, either previously uncropped ground or land planted to rice, peanuts, maize or potatoes. A new mill has been commissioned west of Mareeba, with its initial crush during the 1998 season. Other industry Issues Other issues for horticulture in north Queensland include: 1. Poor returns to mango growers and inadequate research dollars from that industry, 2. The ‘selection’ of five exotic tropical tree crops with market potential - abiu, durian, longan, mangosteen and rambutan; 3. The impending loss or restricted use of endosulfan, which has significance for many crops, particularly those affected by fruitspotting bugs; 4. The establishment of a new post-harvest facility in Cairns, following the Asian Papaya Fly incursion; and 5. The formation of the Queensland Horticulture Institute in 1997 as a business unit of DPI. New or Resurgent Pests Apart from the exotic pests outlined in Table 1, a number of other insects and mites have caused concern for horticultural production in north Queensland since 1991. The silverleaf whitefly, B. tabaci type B, has been increasing in prominence, particularly in the vegetable and melon crops of the Burdekin Irrigation Area. The Queensland Horticulture Institute has recently employed an entomologist to examine the problem, including the insecticide resistance issue. Spider mites, and particularly the two-spotted mite, Tetranychus urticae Koch, have progressed from being a seasonal or insecticide-induced problem to a perennial and severe pest for some papaw growers. The sugarcane weevil borer, R. obscurus, has continued to trouble the palm industry, and is again a concern for sugarcane growers with the wide adoption of green cane harvesting. A new weevil pest of macadamias, an unknown Sigastus sp., appeared in Atherton Tableland crops during 1994/95. Data collected so far suggest that this weevil can cause significant nut losses in orchards, which are organic or minimally treated with insecticide. A moth thought to belong to the genus Cryptophlebia has emerged as a serious threat to the avocado industry on the Atherton Tableland. The female lays its eggs on the fruit surface, and the emerging larvae rapidly bore through the epidermis. The insect appears in crops from late January to early March and particularly threatens varieties such as Hass. Research on this pest has become a high priority for the local industry. Another pest not previously known to north Queensland horticulture is the small mirid, Taylorilygus nr. nebulosus (Malipatil pers. comm.). It feeds on developing avocado and longan fruit, in the former causing lumpy fruit on maturity. The exotic pest incursions into north Queensland since 1991 are given for both Torres Strait and the mainland in Table 1. Following the detection of the Asian Papaya fruit fly in Cairns in 1995, spiralling whitefly, Aleurodicus dispersus Russell, was located in the city in 1998. Its containment is being undertaken by localised treatment, restricted movement of host plant material and distribution of the wasp parasitoid, Encarsia. The detection of both mango 13 leafhopper and red-banded mango caterpillar in the Peninsula/Torres Strait areas is a concern for the mango industry coming soon after their Asian Papaya fruit fly experience. Table 1. Exotic pest incursions into north Queensland since 1991 Torres Strait Mainland Bactrocera papayae, B. trivialis, B. cucurbitae (regularly since 1993) Bactrocera papayae (Cairns - Oct.1995) Noorda albizonalis (Saibai - 1989/90) (Dauan - 1996/97; ?Moa-1997) Aleurodicus dispersus (Seisia - Mar. 1995) (Cairns - Mar. 1998) Apis cerana, Varroa jacobsoni (Saibai,Boigu,Dauan - 1992/3) Idioscopus nitidulus (near Weipa - Sept. 1997) Some Issues Facing Northern Entomologists 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. The continued emergence of new pests; The loss or restricted use of valuable insecticides due to the ongoing review process; Increasing insecticide resistance, exacerbated by the loss of certain groups of compounds; Lack of new chemicals to replace old ones; Constraints on the import or release of exotic beneficials; Limits to the development and availability of new technologies; Ongoing expansion in the range of crops and the associated pest problems; Consumer demands for cleaner, greener produce; Market access/quarantine requirements; and Increasing competition for research dollars. Acknowledgements The Queensland Horticulture Institute has authorised the publication of this paper. References Anon. (1998). Directions: A program for agribusiness on the Tablelands. Tablelands Advertiser, Wednesday June 24, 1998. Pp. 21 - 28. Fay, H.A., Drew, R.A. and Lloyd, A.C. (1997). The eradication program for Papaya fruit fly (Bactrocera papayae Drew and Hancock) in north Queensland. Pp. 259-261. In Allwood, A.J. and Drew, R.A.I. (Eds.) Management of Fruit Flies in the Pacific: A regional symposium, Nadi, Fiji 28 - 31 October 1996. 14 BIOLOGICAL CONTROL OF WEEDS IN THE NORTHERN TERRITORY Andrea Wilson Biological Control Section, Weeds Branch Department of Primary Industry and Fisheries GPO Box 990 Darwin NT 0801 Abstract The DPIF has initiated biological control programs against five weeds in conjunction with CSIRO Entomology. To date, 11 biological control agents, nine insects and two pathogens, have been released against the woody weed, Mimosa pigra in the Northern Territory. Of these, seven insects and a pathogen have established. Three insects have been released against the malvaceous shrub, Sida acuta. A chrysomelid beetle has established, is now widespread across the Top End and appears to reduce seed production. Overseas exploration has commenced in a program targeting the shrub Jatropha gossypifolia and plans to begin exploration for agents against the annual herb, Argemone ochroleuca are underway. Work with Hyptis suaveolans is on hold. Introduction Invasive weeds have been recorded in the Northern Territory (NT) since European exploration began. Hytpis suaveolens (L.) Poit for example, was first reported by the ill-fated explorer Ludwig Leichardt at Port Essington in 1845 (Miller and Schultz 1997). Many invasive plant species originate from tropical countries. A large number were introduced for pasture “improvement” or as ornamental garden plants (Smith 1995). The rapid spread of one such plant, Mimosa pigra L., was the stimulus for the Weeds Branch, DPIF to initiate its first biological control program against the weed in 1979 (van Rangelrooy 1994). Mimosa pigra has the same origin (tropical America) as a number of other NT weeds. Other projects were commenced in conjunction with the M. pigra program and in collaboration with CSIRO Entomology. I shall discuss the progress to date of biological control of M. pigra and Sida acuta Burman F. and other current programs and studies undertaken by the Biological Control Unit of the Weeds Branch, DPIF. Mimosa pigra L. Mimosa pigra, which originated in Central America, has now spread throughout the tropics and threatens all wetlands of tropical Australia. The plant is thought to have been introduced into Australia through the Darwin Botanic Gardens in the late nineteenth century as a curiosity plant. By the 1950s M. pigra had spread to the Adelaide River township, at the end of the 1960s, the Marrakai crossing and in 1975 had reached the Arnhem Highway bridge (Miller and Lonsdale 1987). From there the weed spread rapidly over the flood plains aided by floodwater and overgrazing by buffalo (Lonsdale et al. 1988). At the present time the plant covers at least 80 000 hectares of the Northern Territory (Lonsdale 1992) and is arguably the worst weed of the Top End. Mimosa pigra now occurs from the Fitzmaurice River in the west to the Phelps River in the east. Scattered infestations at Maningrida, Arafura Swamp, Stray Creek on Jindare station, Scott Creek west of Katherine and on North Perron Island are being controlled using herbicides with a view to eradication. Larger infestations are chemically and mechanically treated to prevent further spread and allow economic or cultural activity. The main problem is the ability of M. pigra, in the introduced range, to produce a large number of seeds. For example the seed bank at a site in the lower Adelaide River in the NT was found to be 10,000 seeds/m2, two orders of magnitude higher than in Mexico (Lonsdale et al. 1988). 15 Thus a primary objective of the biological control program is to reduce viable seed production and hence the ability of M. pigra to colonise new areas. Mimosa pigra is a resilient plant. Eleven biological control agents, nine insects and two pathogens, have been released against the weed (Table 1). All insects, except for Chalcodermus serripes and Sibinia fastigiata, have established. Table 1. Biological control agents against Mimosa pigra as of June 1998 AGENT FAMILY RELEASE * Acanthoscelides puniceus Johnson Bruchidae April 1983 A. quadridentatus (Schaeffer) Bruchidae April 1983 Chlamisus mimosae Karren Chrysomelidae November 1985 Neurostrota gunniella (Busck) Gracillariidae February 1989 Carmenta mimosa Eichlin and Passoa Sesiidae July 1989 Coelocephalapion aculeatum Fall Apionidae January 1992 Coelocephalopion pigrae Kissinger Apionidae May 1994 Phloeospora mimosae-pigrae Coelomycetes March 1995 Diabole cubensis Urenidales March 1996 Chalcodermus serripes Fahreus Curculionidae April 1996 Sibinia fastigiata Clark Curculionidae December 1997 * First released in the Northern Territory The bruchids, Acanthoscelides puniceus and A. quadridentatus were the first agents released (Table 1). Both insects established quickly and spread rapidly. However, by 1990, seven years after the initial release, less than one per cent of seed was consumed by these beetles prior to dispersal (Wilson and Flanagan 1991). Another seven years later in a survey of biological control agents of M. pigra conducted in September 1997 the overall pre-dispersal consumption by the bruchids had risen to four per cent. Up to 25% pre-dispersal seed destruction was recorded at one site (unpublished data). The next insect to be released has had little impact. Chlamisus mimosae is a Brazilian stemgrazing chrysomelid. The insect is sedentary, has established only on the Finniss River and not spread far from release sites. In 1989 two Mexican moths were released, the sesiid Carmenta mimosa and the gracillarid Neurostrota gunniella (Table 1). Early instars of N. gunniella mine leaf pinnules while the older larvae tunnel into stems and tips (Davis et al. 1991). Predictions have been made of a reduction in seed production by up to 60% through N. gunniella activity. CSIRO Entomology in Darwin is currently quantifying the insect’s impact. Today this moth is practically ubiquitous. Nearly all infestations of M. pigra are colonised by the moth (unpublished data). The larvae of C. mimosa tunnel into larger M. pigra stems. Edge plants are commonly heavily infested. This moth causes considerable damage to plants in high numbers. The life cycle of C. mimosa is longer than that of N. gunniella (12 weeks compared with 4.5 weeks at 25° C) (Forno et al. 1992). While this insect has established over more than 220 km2 (unpublished data) numbers of C. mimosa have fallen over the last two wet seasons. Record rainfall and subsequent high water levels appear to have caused larvae to drown and populations crash (pers. obs.). Larvae are still being 16 harvested from plants grown at Berrimah Agricultural Research Centre (BARC) and are currently being redistributed to sites on the Mary and Daly Rivers. Impact studies to quantify C. mimosa effect on plant growth and seed production are underway. Coelocephalapion aculeatum was the first of the apionid weevils released (Table 1). Both adult and larvae consume buds and flowers of M. pigra, exclusively (Forno et al. 1994). The insect established but has not been found away from release sites (M. Paskins pers. comm.). The second apionid, C. pigrae, was released two years later. Within three years the insect was found at Elizabeth Downs station, at least 80 km from the nearest release site (pers. obs.) and now appears as ubiquitous as N. gunniella. Their success may be due in part to their diet as adult of C. pigrae consume both flowers and leaves (Heard 1996) thus being able to survive the dry season when flowering occurs infrequently. In the plant’s native range, two green seed feeding weevils are estimated to consume more than 95% of green seeds of M. pigra (T. Heard pers. comm.). In 1996 the mature green-seed feeder Chalcodermus serripes was first released in the NT. The female lays on green seeds greater than 4 mm in length. This weevil appears to be univoltine, entering what appears to be an obligate diapause at the prepupal stage and thus difficult to rear in the laboratory (Heard et al. 1997). Direct release permits were sought and C. serripes are released into the field after stringent screening in quarantine. Establishment on the Finniss River has not been confirmed but evidence of insect activity persists at the release site. The second green seed feeder, Sibinia fastigiata, was first released in December 1997. Sibinia fastigiata oviposit on seeds less than 3 mm long, thus complementing the impact of C. serripes. Adult S. fastigiata feed only on nectar and pollen of M. pigra and a closely related plant (Heard et al. 1997). Due to difficulty rearing these insects in the laboratory - concomitant flowering seeds of a suitable size for oviposition and an ovipositing female are required (Heard et al. 1997). This weevil is also released directly into the field. As the insect is multivoltine and adult emergence is triggered by rain (T. Heard pers. comm.) which also initiates flower production, this weevil appears to be well adapted to northern Australian conditions. At the present time the Weeds Branch is continuing to release the seed feeders and redistributing C. mimosa. CSIRO Entomology in Brisbane is currently host-testing lepidopteran leaf tiers. A comprehensive survey of the distribution of biological control agents on M. pigra in the NT is almost complete. Present work also centres on experiments to determine the most cost-effective control method of the weed and studies of the impact of N. gunniella, C. pigrae and C. mimosa on M. pigra are underway. Sida acuta Burman f. Spiny head sida is native to central America and thought to have been introduced into the NT by Chinese miners for use as brooms in the nineteenth century (Pitt 1992). Heavy grazing and the plant’s unpalatability have led to the formation of monospecific stands of S. acuta in overgrazed paddocks, around watercourses and stockyards (Lonsdale et al. 1995). Sida acuta has become a problem throughout the Darwin, Gulf and Victoria River regions of the NT (Pitt 1992). To date, three biological control agents have been released against S. acuta (Table 2). A leaffeeding chrysomelid beetle, Calligrapha pantherina was released on the Finniss River in 1989. Both larvae and adult consume the leaves and flowers of the plant (Forno et al. 1992). The beetle usually appears in destructive numbers in mid to late wet season in coastal areas. Populations sweep through stands of S. acuta, completely defoliating plants. This year they were found on the banks of the Katherine River (before the floods), in the Pine Creek area and on most properties in the Darwin area (pers. obs.). In early 1997, five years after its initial release, C. pantherina was found in high numbers on Mataranka station, 100 km south of Katherine. The manager estimated a return of 70% pasture grass in what was a monoculture of S. acuta after defoliation by the insect the previous wet season. Seed production at one site in the Darwin region was reduced by an order of magnitude after defoliation by the beetle (Lonsdale et al. 1995). 17 There are problems with C. pantherina. Firstly, defoliation often occurs after seeding (pers. obs.). Secondly, C. pantherina occurrence is seasonal especially in the drier areas of the weed’s range. Recovery from dry season population crashes may take several months. And thirdly, the beetle often does not kill larger plants. Other insects were selected to complement and enhance that action of C. pantherina. A stem boring weevil, Eutinobothrus species was initially released in 1994 (Table 2). This insect is currently being described. Larvae feed internally while adults graze on the stem of the plant. Eutinobothrus sp. aestivates during the dry periods, so numbers of the weevil should survive the dry season ready to emerge with the first rains (Day et al. 1995). Table 2. Biological control agents released against Sida acuta INSECT FAMILY RELEASE* Calligrapha pantherina St. I Chrysomelidae September 1989 Eutinobothrus sp. Curculionidae January 1994 E. pilosellus (Boheman) Curculionidae March 1997 * First released in the Northern Territory Observations of interaction between the agents appear promising. At a property near Darwin S. acuta plants infested with the weevil and then grazed by C. pantherina had visibly higher mortality than those defoliated by the chrysomelid beetle alone (pers. obs.). A second weevil, E. pilosellus, was initially released in 1997 (Table 2). Larvae honeycomb the stem while adults graze on leaves leaving characteristic scars. This insect can reproduce in stems of smaller diameter than Eutinobothrus sp. (G. Fichera pers. comm.). Stem integrity is also maintained: Eutinobothrus sp. creates holes where frass is pushed to the exterior of the plant. Thus E. pilosellus appears better adapted for survival in areas prone to flooding. Studies to quantify the impact of weevils are the underway. Jatropha gossypifolia L. Belly-ache bush or Jatropha gossypifolia is a significant pest in India, Jamaica, Brazil, Trinidad, Africa, Papua New Guinea, northern Australia and eastern Indonesia (Pitt 1997). The plant was probably introduced as a garden plant last century. Bellyache bush is a tough perennial plant and forms dense thickets along waterways and degraded land. The plant is spread by seeds and vegetative parts of the plants. Overseas exploration for agents has begun. A cerambycid from Venezuela appears promising. Host testing of the beetle should commence this year (IW Forno, pers. Comm.). Argemone ochroleuca Sweet As with many of the weeds in the NT, Mexican poppy, Argemone ochroleuca, originated in Central America. Now considered as one of the world’s worst weeds, A. ochroleuca occurs throughout the tropics (Wilson 1995). This thistle-like plant was probably introduced into NSW through contamination of grain (Smith 1995). In recent years Mexican poppy has spread rapidly through riverbeds of central Australia, changing the appearance of the country and posing a threat to tourism and the pastoral industry. Overseas exploration for biological control agents should begin within a year. Hyptis suaveolans (L.) Poit. Hyptis or horehound is an aromatic plant that grows to a height of two metres (Miller and Schultz 1997). This plant is said to be the most abundant and widespread weed in northern 18 Australia. Potential biological control agents have been identified but work on the weed has been placed on hold until extra funds become available or work on higher priority weeds is completed. Acknowledgements I thank Grant Flanagan for valuable comments. References Davis DR, Kassulke RC, Gillet JD and Harley KLS. (1991). Systematics, morhpology, biology and host specificity of Neurostrota gunniella (Buesk) (Lepidoptera: Gracillaridae), an agent of biological control of Mimosa pigra in Australia. Entomological Society of Washington 93, 16-44. Day MD, Forno IW, Segura R and Martinez M. (1995). Life cycle and host specificity of Eutinobothrus sp. (Col: Curculionidae) an agent for biological control of Sida acuta (Malvaceae) in the Northern Territory, Australia. Entomophaga 40, 345-355. Forno IW, Heard TA and Day MD. (1994). Host specificity and aspects of the biology of Coelocephalapion aculeatum (Coleoptera: Apionidae), a potential biological control agent of Mimosa pigra. Environmental Entomology 23, 147-153. Forno IW, Kassulke RC and Day MD. (1992). Life-cycle and host testing procedures for Carmenta mimosa Eichlin and Passoa (Lepidoptera: Sesiidae), a biological control agent for Mimosa pigra L. (Mimosaceae) in Australia. Biological Control 1, 309-315. Forno IW, Kassulke RC and Harley KLS. (1992). Host specificity and aspects of the biology of Calligrapha pantherina (Col: Chrysomelidae), a biological control agent of Sida acuta (Malvaceae) and S. rhombifolia in Australia. Entomophaga 37, 409-417. Heard TA and Forno IW. (1996). Host selection and host range of the flower-feeding weevil, Coelocephalapion pigrae, a potential biological control agent of Mimosa pigra. Biological Control 6, 83-95. Heard TA, Forno IW and Burcher JA. (1997). Chalcodermus serripes (Coleoptera: Curculionidae) for biological control of Mimosa pigra: host relations and life cycle. Unpublished report. Heard TA, Segura R, Martinez M and Forno IW. (1997). Biology and host range of the greenseed weevil, Sibinia fastigiata, for biological control of Mimosa pigra. Biocontrol Science and Technology 7, 631-644. Lonsdale WM. (1992). The biology of Mimosa pigra. In: A Guide to the Management of Mimosa pigra (ed KLS Harley) pp.8-32. CSIRO, Canberra. Lonsdale WM, Farrell G and Wilson CG. (1995). Biological control of a tropical weed: a population model and experiment for Sida acuta. Journal of Applied Ecology 32, 391-399. Lonsdale WM, Harley KLS and Gillett JD. (1988). Seed bank dynamics in Mimosa pigra, an invasive tropical shrub. Journal of Applied Ecology 25, 963-976. Miller IL and Lonsdale WM. (1987). Early records of Mimosa pigra in the Northern Territory. Plant Protection Quarterly 2, 140 -142. Miller IL and Schultz GC. (1997). Hyptis or Horehound (Hyptis suaveolans). Agnote No. 477, September. Weeds Branch, Darwin. Department of Primary Industry and Fisheries. Pitt J.L. (1992). Spinyhead sida (Sida acuta). Agnote No. 496, February. Weeds Branch, Darwin. Department of Primary Industry and Fisheries. 19 Pitt J.L. (1997). Bellyache bush (Jatropha gossypifolia). Agnote No. 480, September. Weeds Branch, Darwin. Department of Primary Industry and Fisheries. Smith N.M. (1995). Weeds of Natural Ecosystems. Environment Centre, Darwin. Van Rangelrooy DS. (1994). Biological Control of Mimosa pigra. Agnote No. 594, September. Weeds Branch, Darwin. Department of Primary Industry and Fisheries. Wilson C.G. (1995). Mexican poppy (Argemone ochroleuca). Agnote No. 632, May. Weeds Branch, Darwin. Department of Primary Industry and Fisheries. Wilson C.G and Flanagan (1991). Establishment of Acanthoscelides quadridentatus (Schaeffer) and A. puniceus Johnson (Coleoptera: Bruchidae) on Mimosa pigra in the Northern Territory. Journal of the Australian Entomological Society 30, 279-280. 20 REVIEW OF THE CURRENT COTTON RESEARCH PROGRAM IN THE ORD A. J. Annells1 and G. R. Strickland2 Agriculture Western Australia, PO Box 19, Kununurra Western Australia 6743 2 Agriculture Western Australia, 3 Baron-Hay Court, Perth Western Australia 6151 1 Introduction The cotton research program in the Kimberley commenced in 1993 with a study on the feasibility of growing INGARD® cotton (Strickland and Fitt 1993). The feasibility study found that it should be possible to grow INGARD® but there were some major constraints that had to be investigated if the industry was to be sustainable. The current cotton research program in the Ord River Irrigation Area (ORIA), carried out jointly by Agriculture Western Australia and CSIRO, is investigating these constraints and developing a best management practice for growers. Commercial Cotton Production (1963 - 1974) The only significant period of cotton production in the Kimberley occurred from 1963 to 1974 in the ORIA. During this time the area planted to cotton peaked at 4,795 ha and the highest yield was achieved in 1971 with 4.8 bales/ha which was equal to yields being achieved in eastern Australia at the time. Today, with advances in technology, plant breeding and Australian adapted cotton varieties, the average yield for the eastern States is around 8 bales/ha (The Australian Cotton Grower 1998). The early cotton industry in the ORIA was plagued with quality problems and pest management difficulties, which eventually over-whelmed the industry. Problems associated with the cropping system included: • • • • • • • • wet season cotton production; pest control and crop management delayed by rain; only broad spectrum insecticides were used; insect control was not done on an integrated pest management basis; ratooning practiced on at least 25% of crop; 12 month cropping cycle with other heliothis hosts (e.g. sorghum); harvesting occurred over a protracted period; insecticide resistance developed. INGARD® Cotton Production (1993 →) The feasibility study in 1993 decided that INGARD® cotton could not be grown sustainably in the Kimberley using broad spectrum insecticides alone, or with the management practices of the past (Strickland and Fitt 1993). So, how can we contemplate a cotton industry again? Technological advances have led to: • • • improvements in harvesting, transport and storage of lint prior to ginning; improvements in plant varieties and transgenic plant varieties; new insecticides which are very specific to the target. However, new plant varieties, transgenics (INGARD®) and new generation insecticides are not complete solutions to past problems. Many of these new products require very specific conditions for application, or performance is compromised. Also, results under high pest pressure may be disappointing. In many cases the new products control only 60% of the pest population, meaning that at high densities you may still have above threshold pest pressure. These new technologies are tools that are ideally suited to an integrated management package. 21 Most of the essential ingredients of an integrated pest management package, such as insect development models, economic thresholds, computer based decision support tools, cultural, biological and chemical control methods are already available to the cotton industry. However, it is arguable whether these have ever been put together and used as one package. This is attempted in the Ord research program where the cornerstone of the IPM package being investigated, is the use of INGARD® transgenic cotton varieties to control lepidopteran pests. Further control is being provided by cultural, biological and chemical methods built around this foundation. I will now compare some of the new approaches with past practices. Wet Season vs Dry Season Production One of the most radical features we are trialing in the ORIA cotton production system is winter (dry season) production rather than summer (wet season) production. Temperatures in the Ord during winter are warm enough for cotton growth with maximum temperatures ranging from 30.3 - 38.3oC and minimum temperatures from 14.4 - 23.5oC. However, there is a 5% chance of the minimum temperature falling below 12oC on any one night during the winter months. This represents the critical threshold temperature for cotton growth. Below 12oC cotton stops growing and the plant may be physiologically damaged. By growing cotton through the winter two major pests of the previous cotton industry Spodoptera litura and Pectinophora gossypiella (Richards 1948) are avoided. Heliothis activity is generally less because temperatures are lower than in summer so the insect’s life cycles are lengthened. There is also some improvement in insecticide efficacy brought about by cooler temperatures. Dry season cotton production has restricted growers in the ORIA to very narrow planting and harvesting windows in order to avoid the wet season. The initial feasibility study indicated that cotton could be planted from March to June. Such an extended planting window can favour pest insects because it provides a long period of time over which favourable food is available. Time of sowing trials carried out over a number of years indicated that mid-March to mid-April is the optimum period for planting (Yeates and Constable 1998). Yields are maximised, compensating for any quality discount due to short fibre length and flowering and boll development occur during the months of May to July when pest numbers are low. By the time pest numbers build up the crop has become a largely unfavourable food site. Mid-March to mid-April sowing dates also ensure that harvesting is completed by late October so that rain does not cause weathering of fibre or destroy the crop (Yeates and Constable 1998). Dry season cotton production has created some interesting agronomic problems that are being investigated by CSIRO. The major concerns are short fibre length, which is influenced by temperature, and excessive early vegetative growth, which causes rankness. Early rankness leads to problems with spray penetration and insect control later in the season. Conventional Cotton versus INGARD® Cotton In 1996 conventional cotton and INGARD® were grown with the same IPM program and the numbers of heliothis larvae, sprays and yields were compared. The cotton was grown with lucerne strips planted every 300 m. The lucerne acts as a trap/companion crop. The beneficial food spray Envirofeast™ was also applied. Conventional chemicals were used to control pests based on results obtained from crop scouting and EntomoLOGIC™ recommendations. More heliothis larvae were found on the conventional cotton than on the INGARD® for the entire season (Figure 1). 22 20 18 Mean Larvae Per Metre 16 14 Conventional Cotton 12 10 8 6 4 Ingard Cotton 2 Oct 7 Sep 30 Sep 23 Sep 16 Sep 9 Sep 2 Aug 26 Aug 19 Aug 12 Aug 5 Jul 29 Jul 22 Jul 15 Jul 8 Jul 1 Jun 24 Jun 17 Jun 10 Jun 3 May 27 May 20 May 13 0 Figure 1. The mean number of Heliothis larvae found on conventional cotton and INGARD® cotton grown with the same IPM program in 1996 in the ORIA The conventional cotton therefore required more sprays than the INGARD® in order to produce a similar yield (Table 1). Table 1. The total number of insecticide sprays applied for two major pests, mirids and heliothis, to Conventional and INGARD® cotton grown with the same IPM program and the resultant yields at harvesting Treatment Mirid Sprays Heliothis Sprays Yield (bales/Ha) L23conv + IPM 3.0* 7.5 7.02 L23i + IPM 1.25 1.75 7.43 * Some mirid sprays were also for rough bollworm INGARD® Alone versus INGARD® with IPM Many ask why are we bothering with IPM when other growers throughout the world are growing INGARD® successfully without it. There are a number of important issues that particularly affect the Kimberley and have not been fully addressed in other areas yet. These include: • • • resistance to INGARD® may develop unless pre-emptive strategies are developed; there is usually a lack of prey (heliothis larvae) in INGARD® crops and this may discourages predators; and INGARD® does not last all season and late season control of heliothis is important for both INGARD® resistance management and insecticide resistance management. Over the past two years INGARD® cotton has either been grown by itself or with lucerne and IPM. All the crops were scouted for pests and sprayed on entomoLOGIC thresholds. Insecticides used on each crop varied from year to year but first option chemicals were always ones that were specific to the target and soft on beneficial insects, particularly in the IPM treatments. We have found that in the ORIA there is a point reached late in the crop cycle where we have no option but to abandon IPM and use knockdown insecticides. In both 1996 and 1997 there were more heliothis larvae found on the INGARD® cotton grown alone than on the INGARD® cotton grown with IPM. (Figure 2). The INGARD® cotton grown 23 alone therefore required more insecticide sprays than the INGARD® with IPM in order to produce a similar yield (Table 2). 2.0 Mean Larvae Per Metre a 1996 1.5 1.0 0.5 0.0 5/1/96 5/22/96 6/12/96 7/3/96 2.0 Mean Larvae Per Metre b 7/24/96 8/14/96 9/4/96 9/25/96 1997 1.5 1.0 0.5 0.0 3/25/97 4/15/97 5/6/97 5/27/97 6/17/97 7/8/97 7/29/97 8/19/97 9/9/97 9/30/97 10/21/97 Ingard Alone Ingard with Lucerne Figure 2. Mean number of Heliothis larvae per metre on INGARD cotton grown alone or INGARD cotton grown with IPM in a. 1996 and b. 1997 in the ORIA Table 2. The total number of insecticide sprays applied for two major pests, mirids and heliothis, to INGARD® cotton grown alone and INGARD® cotton grown with IPM program and the resultant yields at harvesting Treatment Mirid Spray Heliothis Spray Total Spray Yield (bales/Ha) L23i alone 1.75 1.75 3.5 7.35 L23i + IPM 1.25 1.75 3.0 8.65 L23i alone 2.5 2.75 5.75 6.61 L23i + IPM 1.25 1.75 3.25 6.82 1996 1997 Once every week in 1997, 100 Helicoverpa spp eggs were collected from the leaves of INGARD® cotton grown alone or from INGARD® cotton grown with IPM and the number of Trichogramma which, emerged from the eggs was recorded. Trichogramma is one of the most important natural control agents of heliothis in the ORIA. There was very little difference between the INGARD® alone and the INGARD® with IPM in the number of heliothis eggs parasitised by Trichogramma (Figure 3). 24 1.0 Proportion Parasitism 0.8 0.6 0.4 0.2 0.0 4/15/97 4/29/97 5/13/97 5/27/97 6/10/97 6/24/97 Ingard Alone 7/8/97 7/22/97 8/5/97 Ingard with IPM Figure 3. The proportion of Heliothis eggs parasitised by Trichogramma in INGARD® cotton grown alone and INGARD® cotton grown with IPM in the ORIA in 1997 In 1997 Devac samples were taken regularly from the INGARD® cotton grown alone and also from the INGARD® cotton grown with IPM. The insects were sorted to species and numbers per metre were counted. In 1997 beneficial insects of heliothis included ladybird beetles (Coccinella transversalis), green and brown lacewings, predatory Hemiptera (Nabis sp., Geocoris sp., Deraeocoris signatus, Orius sp.) and spiders. More beneficial insects were found on the INGARD® with IPM than on the INGARD® grown alone, particularly late in the season (Figure 4). 15 Mean Beneficials per Metre 12 9 6 3 Ingard Alone Sep 26 Sep 19 Sep 12 Sep 5 Aug 29 Aug 22 Aug 8 Aug 15 Aug 1 Jul 25 Jul 18 Jul 11 Jul 4 Jun 27 Jun 20 Jun 6 Jun 13 May 30 May 23 May 9 May 16 May 2 Apr 25 Apr 18 Apr 11 0 Ingard with Trap Crop Figure 4. Mean number of total beneficial insects (excluding Trichogramma sp.) found in INGARD® cotton grown alone and INGARD® cotton grown with IPM in the ORIA in 1997 Probably the most frustrating part of any IPM program, and one that is proving common in the ORIA, is that there is always another pest which must be fitted into the system. In 1997 green mirids, Creontiades dilutus, were a serious problem. Endosulfan was used to control mirids because it has low toxicity to beneficial insects. Unfortunately it was not very effective because the mirids were high mobility and endosulfan had low persistence. A number of alternative chemicals have been trialled and we believe we have found an alternative to endosulfan. In 1998 Monolepta australis, the red-shouldered leaf beetle was a serious threat to seedling stands. The adult beetles strip the plants of leaves and terminals in minutes, killing the seedlings. We are currently treating infestations with spot sprays of broad-spectrum knockdowns before they become too large and widespread. Unfortunately monitoring for an infestation is a problem because the aggregations develop rapidly and a large amount of 25 damage can occur in a short space of time. Also, if the aggregations are in the centre of the crop they are hard to see until serious damage occurs. Work needs to be done on a more efficient monitoring process and also on control of the beetles. Over the next few years Agriculture Western Australia plans to continue developing and refining the cotton IPM system. Validation work on many of the spray thresholds in the entomoLOGIC program will be carried out for the ORIA. Agriculture Western Australia has been generously supported in its collaborative research with CSIRO and particularly acknowledges the support of the growers involved with the IPM studies, CRDC, Monsanto, CSD, Deltapine, Colly Cotton and CSIRO. References Richards, K.T. (1968). A study of the insect pest complex of the Ord River Irrigation Area. M.Sc. Thesis, University of Western Australia. Strickland, G.R., Addison, S.J., and Annells, A.J. (1998) IPM and INGARD® in the Ord. The Australian Cotton Grower 19 48-53. Strickland, G.R., and Fitt, G.P. (1993) Assessing the potential for sustainable cotton production in the Kimberley: Pest management study. Unpublished (available from Agriculture Western Australia). The Australian Cotton Grower (1998) Cotton Yearbook. The Australian Cotton Grower, Toowoomba. Yeates, S.J. and Constable, G.A. (1998) Ord update: Refining the production system. The Australian Cotton Grower 19 45-47. 26 ERADICATION OF THE ORCHID WEEVIL ORCHIDOPHILUS ATERRIMUS (WATERHOUSE) (COLEOPTERA: CURCULIONIDAE) FROM THE NT ESC SMITH and MJ NEAL Entomology Branch Department of Primary Industry and Fisheries GPO Box 990 Darwin NT 0801 Abstract The orchid weevil is a serious pest of Dendrobium orchids in many parts of the world. It was probably introduced to the Northern Territory (NT) on Phalaenopsis orchids purchased from Cairns, north Queensland in 1986. Nursery surveys during 1991 indicated that it had a very limited distribution in the NT, and an eradication program was mounted over a 10 month period in 1992-93. Regular inspections over the subsequent 26 months failed to record the presence of the pest. The quarantine area was revoked and the orchid weevil declared eradicated from the NT in October 1995. Introduction The orchid weevil is a serious pest of Dendrobium orchids in many parts of the world. It is probably native to Malaysia and Thailand but now occurs in the Philippines, Indonesia, Japan, Singapore, Hawaii and north Queensland. It was probably introduced to the NT on Phalaenopsis orchids purchased from Cairns in 1986. Surveys of nurseries in the Top End during 1991 indicated that it had a very limited distribution in the NT. After confirming that only a single property was infested, an eradication program was mounted over a 10 month period in 1992-93. Regular inspections over the subsequent 26 months failed to record the presence of the pest. Then the orchid weevil was declared eradicated from the NT in October 1995. This paper briefly describes the life cycle, pest status, host range and the damage caused by the weevil based on overseas information and outlines the eradication of the pest from the NT. Life Cycle and Damage Caused by the Weevil Adult weevils are small black beetles with a typical, long weevil "snout". These pests are about 4-5 mm long and feed on the succulent flowers, leaves, pseudo-bulbs and exposed roots of orchids. The soft apical growth is preferred. They are long lived, surviving over a year under shade house conditions in Hawaii. Eggs are laid singly inside feeding cavities in the pseudo-bulbs or at the base of developing leaves, or inside leaves in the laboratory (Mau 1983). The larvae (grubs) do not have legs and feed soon after hatching, mainly inside the pseudo-bulbs. Larval development occurs within galleries inside the pseudo-bulbs. The gallery is usually orientated downward and filled with frass (excreta) to a short distance behind the larvae. The duration of larval development is from three to more than five months and is longer in old, harder pseudo-bulbs. There are five, possibly six, larval stages. The damaged pseudo-bulbs are not usually killed but often cease growing or producing flowers. Pupation occurs at the end of the gallery and takes between 13 and 18 days. Newly emerged adults do not feed for ten days and remain in the pseudo-bulbs for two to three weeks before emerging through a circular hole 2-3 mm across. They are initially light-dark brown but turn black after three to four days. 27 Host Range and Overseas Control Methods In Hawaii, these pests are commonly found on Dendrobium, Vanda and Phalaenopsis but have also been collected from Renanthera, Saccolobium, Cymbidium and Spathoglottis (Mau 1983). No natural controls are known for this pest but several insecticides (e.g. acephate, bendiocarb, chlorpyriphos, methyl parathion) have shown good control when applied at three-weekly intervals (Hara and Mau 1986). In Hawaii, it was found that after insecticides were used, flower production increased and the incidence of white streaking (not due to thrips feeding) was reduced (Hara and Mau 1986). Occurrence in the NT The pest was probably introduced to the NT on Phalaenopsis orchids purchased from Cairns in 1986. When noticed in 1988, the grower attempted to control the weevil with weekly methidathion sprays, which after six months had given good control. However, low numbers persisted. In April 1992, officers from the Entomology Branch inspected the property and noted the presence of O. aterrimus adults and moderate damage to Dendrobiums and Vanda orchids. Since recent surveys of nurseries in the Top End had failed to record the orchid weevil on other properties we suspected that it had a very limited distribution in the NT. The Entomology Branch sought information from both commercial and other orchid growers on the occurrence of this pest. Method During a five month period in 1992, letters were sent to the four orchid societies in the NT, requesting information on the distribution of the orchid weevil and Entomology staff examined any suspected sighting to confirm the presence of weevils. The few responses and lack of positive sightings, indicated that it was likely that a single infestation occurred in the NT and was confined to a relatively isolated property. A decision to attempt eradication of this pest was made and O. aterrimus was declared a “pest” and a “notifiable” pest in September 1992. The affected property was placed under a gazetted quarantine area. An eradication program was instituted in October 1992 which involved strict hygiene conditions: the complete immersion of all susceptible plant materials in chemical dips; subsequent regular insecticide applications to all plants in and around the shadehouse; and regular plant and property inspections. Restrictions on the removal of any plant material from the property were also imposed. Results At the beginning of the program, any damaged orchids or excess plant material was sorted, removed and burnt and the previously overgrown shadehouse re-arranged to provide easy access for chemical treatments and inspections. All Dendrobium material was dipped in 400 ppm dimethoate solution whilst all other orchids and plant species were sprayed to run off with chlorpyrifos. Over the 10 month period to August 1993, 11 insecticide treatments were applied and orchids and other material inside the shadehouse or surrounding areas thoroughly checked for the presence of orchid weevils or plant damage. No fresh weevil damage was observed. Thorough plant inspections were carried out on 10 occasions over the next 26 months. Although some damage to plants was noted and adults, larvae or pupae of the orchid beetle Stethypachys formosa Baly (Coleoptera: Chrysomelidae) were seen during this inspection period, no orchid weevils were detected nor any damage attributed to this pest. It was noticeable that, after the weevil was controlled, similar increases in flower production to that recorded overseas occurred on the treated property. 28 In October 1995, after DPIF entomologists were satisfied that eradication had been achieved and no weevils were sighted over the previous three years, the orchid weevil quarantine area was revoked. Current Situation No further detection of O. aterrimus has been made (to May 1998) but the weevil is still gazetted as a pest and a notifiable pest under the Plant Disease Control Act. This will enable swift action under the Act if any weevils are detected. Although the commercial and amateur orchid growers in the NT are generally alert to the risk of accidentally importing the orchid weevil on plant material from Queensland, the risk to the NT remains. References Mau, DFL (1983). Development of the orchid weevil, Orchidophilus aterrimus (Waterhouse). Proc. Hawaiian Ento. Soc., 24(2 and 3): 293-297. Hara, AH and Mau, DFL (1986). The orchid weevil, Orchidophilus aterrimus (Waterhouse): Insecticidal control and effect on vanda orchid production. Proc. Hawaiian Ento. Soc., 24(1): 7175). 29 A REVIEW OF SEXAVA RESEARCH AND CONTROL METHODS IN PAPUA NEW GUINEA G.R.Young Entomology Branch Department of Primary Industry and Fisheries GPO Box 990 DARWIN NT 0801 Abstract The tettigoniid tribe Sexavini contains the genera Segestes, Segestidia and Sexava. Some species, within these genera, cause severe defoliation of bananas, coconuts, oil palm and Pandanus spp. in Papua New Guinea (PNG). The tettigoniids are sometimes called treehoppers but are usually known in PNG by the common name sexava. Sexava adults and nymphs feed on the foliage of the host plant during the night and shelter on the undersides of leaves and in the spear of the palm during the day. Mature females descend from the palm crown at night and lay their eggs in the soil, before returning to the crown. Each female lays 30 to 40 eggs. The number of nymphal instars varies from five to seven and the duration of each instar ranges from 10 to 20 days. Adults live up to 110 days in captivity and probably longer in the field. The development of the eggs of two species of sexava shows an initial egg diapause. Diapause enables the eggs to remain viable during relatively dry conditions. When rain follows a dry spell, the increased soil moisture causes a synchronised egg hatch, giving rise to discrete generations. The sudden increase in the population over one generation, results in the defoliation of palms and consequent loss of production. Most methods of cultural, mechanical and chemical control are either impractical or uneconomic, although trunk injection using monocrotophos is still practised successfully on oil palm. For these reasons much research over the past 55 years has concentrated on biological control. The periodic release of the egg parasites Leefmansia bicolor and Doirania leefmansi to control sexava outbreaks has been a standard method of biological control since before World War II. However, there is evidence to suggest that egg mortality has a negligible effect on sexava populations. Stichotrema dallatorreanum is a strepsipteran parasite of sexava nymphs and adults. Female S. dallatorreanum attack sexava while males parasitize ants. S. dallatorreanum appears to control mainland sexava populations by reducing the fecundity of the host. The life history of the male parasite is incompletely known, which has resulted in the failure of attempts to introduce the parasite in areas where it does not occur naturally. Key words: Segestes, Segestidia, Sexava, coconuts, oil palm, biological control, 1935 to 1990 Introduction The tettigoniid tribe Sexavini contains the genera Segestes, Segestidia and Sexava (Rentz 1996). Some members of these genera cause severe defoliation of bananas, coconuts, oil palm and Pandanus spp. in PNG. The tettigoniids are sometimes called treehoppers but are usually known in PNG by the common name sexava. Most sexava species have green and brown colour forms with an adult body length varying from 45 mm to 60 mm and the antennae many times longer than the body. Macropterpus and micropterous males are known from S. uniformis (Froggatt and O’Connor 1940). Coconuts (copra) are PNG's fourth most important export crop behind coffee, oil palm and cacao. Oil palm is an important cash crop in the West New Britain and Oro provinces. Bananas are an important secondary food over most of the country and a staple food in areas of low rainfall, e.g. the Markham valley, Morobe province. The fruit of Pandanus julianetti is a 31 seasonally important part of the diet in highland villages, while Pandanus conoides and some wild Pandanus spp. are dietary supplements (Bourke 1996). Research into sexava started in 1929 and the Sexava Research Station was set up in the Manus province during 1932 (Froggatt 1935). The station was moved to New Hanover, then to Arawe, New Britain and finally back to Manus prior to 1940 (Froggatt 1935; Froggatt and O’Connor 1940). After World War II sexava research continued at the Lowlands Agricultural Experiment Station, Keravat, East New Britain, and Bubia Agriculture Research Centre, Morobe province (J. Ardley; G. Dunn; C. Perry; personal communication; Bailey et al.1977; Young, 1985, 1987a and 1987b). Sexava outbreaks on oil palm during the late 1970s prompted the start of research on sexava at Bialla, West New Britain province (S. Embupa personal communication). The identity of sexava species remained confused until F. Willemse (1977, 1979) revised the classification of the Sexavini from the Melanesian region. As a result of the confusion work, prior to World War II, was often carried out on more than one species in the belief that the data were being gathered from a single species, e.g. Froggatt (1935) and O’Connor (1937). The purpose of this review is to give an overview of sexava research and control methods, collating data from published papers, extension bulletins and unpublished reports, over the period from 1935 to 1990. Some suggestions are made as to future sexava research with particular reference to biological control. Pest Species and Host Plants Ten species of the Sexavini tribe have been recorded attacking economically important plants (table 1). Most species also feed on a wide variety of native and exotic palms in addition to the economic species (Froggatt and O’Connor 1940; Young (1985). 32 Table 1. Host plant and locality of economically important sexava species PEST SPECIES HOST PLANT LOCALITY REFERENCE Sexava nubila (Stål) coconut (Cocos nucifera) East Sepik Province Young 1987a Segestes decoratus (Redtenbacher) Coconut Mainland PNG, Siassi Island and West New Britian Young 1987a Segestes cornelii F. (Willemse) karuka (Pandanus julianettii) Enga Province DAL records Segestidia leefmansi (C. Willemse) Coconut Lihir Islands, New Hanover and New Ireland Young 1987a Segestidia rufipalpis (C. Willemse) Coconut Misima Island Young 1987a Segestidia uniformis (C. Willemse) Coconut Admiralty Islands, Hermit Islands and Wuvulu Island Young 1987a Segestidia defoliaria defoliaria (Uvarov) Coconut East New Britian, West New Britian, Young 1987a oil palm (Elaeis guineensis) New Ireland and Lihir Islands West New Britian C. Perry pers. comm. Coconut Mainland PNG and Kar Kar Island Oro Province Young 1987a oil palm bananas (Musa spp.) Eastern Highlands Province R. Prior pers. comm. J. Barrett pers. comm. manila hemp (Musa textilis) Morobe Province J. Ardley pers. comm. Segestidia montana F. (Willemse) bananas Eastern Highlands Province J. Barrett pers. comm. Segestidia marmorata (F. Willemse) manila hemp Morobe Province J. Ardley pers. comm. unidentified sexava species Karuka and Pandanus spp. Eastern Highlands and Morobe provinces G.R. Young (unpublished data) Segestidia novaeguineae (Branscik) 33 Life History and Biology Young (1981) has given a generalised life history of sexava, while Froggatt and O’Connor (1940) and Young (1985) have given detailed accounts for S. uniformis and S. decoratus respectively. Coconut palm was the host plant in these studies. The account below is a summary of the work of Froggatt and O’Connor(1940) and Young (1985). Adults and nymphs feed on palm foliage mainly at night and shelter either in the spear of the palm or on the underside of leaflets during the day. S. decoratus, adults and nymphs, prefer to feed on mature coconut fronds while avoiding the young fronds (Young, 1984). Similar observations were made in New Britain by Froggatt and O’Connor (1940). They stated that the fuzz along the edges of leaflets from young fronds apparently clogs the mouthparts of sexava (Froggatt and O’Connor 1940). Mature female sexava descend from the palm crown during the night and lay their eggs in the soil. Sometimes eggs are laid in the butts of old fronds and the roots of epiphytes on the palm trunk. Each female lays 30 to 40 eggs. Females continue to mature eggs over their adult life and as a result eggs are laid over an extended period. The eggs hatch in 50 to 60 days. Within 48 hours of hatching first instar nymphs climb the palm trunk to the crown. The number of nymphal instars varies from five to seven, while the duration of each instar ranges from 10 to 20 days. Adults live up to 110 days in captivity and possibly longer in the field. Mating occurs 10 to 12 days after adult emergence. During copulation the male transfers a spermatophore to the female’s external genitalia. Oviposition starts up to 31 days after mating. Mating occurs several times over the female’s life. P. Room (pers. comm.) found Froggatt and O’Connor’s life history study of S. uniformis to be accurate. Young’s study on S. decoratus indicates the need for further work to more accurately determine the pre-mating and mating to oviposition periods of adult females. There are no published life history studies of any of the other pest species. Adult sexava have poorly developed powers of flight being only able to stay aloft for a distance of 10 metres (Froggatt and O’Connor 1940). Conversely, both nymphs and adults are agile runners and jumpers (Young 1984). Nymphs and adults prefer a high humidity and tend to avoid direct sunlight (Young 1984). This explains why Young (1985) found 98 percent of adults and late instar nymphs of S. decoratus sheltering at the base of fronds adjacent to the spear of coconut palms, at noon - a behavioural characteristic which has ramifications for any sampling program. Froggatt and O’Connor (1940) remarked on the capacity of sexava eggs to withstand prolonged dry conditions. Young (1985) found that the eggs of S. decoratus could survive for at least 75 days at relative humidities in excess of 70%. When S. decoratus eggs were incubated on saturated sphagnum moss and weighed over the duration of their development, water uptake by the eggs showed four separate phases (Young 1990). An initial short water uptake lasting one day, followed by an obligatory quiescent phase of 16 days of no water uptake. A second water uptake phase lasted 38 days during which time the eggs made a weight gain of 116 percent. There was a slight decrease in weight prior to hatching. B. Gorrick (pers. comm.) found that the embryonic development of S. defoliaria did not start until 15 days after oviposition. These findings suggest that the obligatory quiescent phase may be common to many sexava species. The phase is more correctly termed, an initial diapause (Hartley 1990). Fifteen days is the minimum period for initial diapause, however the phase can be extended for several months and the eggs still remain viable (Young 1985). Froggatt and O’Connor (1940) kept eggs in very dry soil for 110 days and then moistened the soil. Within ten days of adding water 40 percent of the eggs had hatched and produced normal nymphs. Young (1985) found that eggs held in saturated soil developed normally, but hatching was inhibited. When the soil was allowed to dry out hatching proceeded normally. The possibility of an optional final embryonic diapause cannot be eliminated, although C. Perry (pers. comm.) was unable to repeat Froggatt and O’Connor’s results (Young, 1985). 34 Losses and Economic Thresholds Little work has been done to either estimate losses due to sexava damage or to determine economic thresholds. Severe defoliation of coconut palms by sexava induces premature shedding of nuts. Froggatt and O’Connor (1940) stated that a defoliated coconut may take up to two years to resume normal production. Bailey et al. (1977) carried out a defoliation trial on coconuts. They found that yield of nuts nine months after the start of defoliation, decreased linearly with increasing defoliation. At 70 percent defoliation, the palms took 17 months to return to normal production after the cessation of defoliation. They concluded that defoliation in excess of 40 percent would affect yield. To obtain feeding rates over all stages of sexava, C. Perry (pers. comm.) conducted a feeding trial with S. decoratus females fed on coconut leaflets. He found that the rates varied from 1.4 cm2 of leaf/day for first instar to 47.7 cm2/day for adults (Table 2). Table 2. Instar duration, leaf area consumed and mean daily feeding rates from three laboratory reared females of S. decoratus Instar Duration (days) Total leaf area consumed (cm2) Mean daily feeding rate, (cm2)/day 1 13.3 14.5 1.4 2 13.3 18.9 1.4 3 14.3 52.0 3.6 4 15.0 113.1 7.5 5 14.0 200.3 14.3 6 16.7 314.0 18.8 7 20.0 472.0 23.6 Adult 71.3 3398.9 47.7 Ecology and Outbreaks Froggatt and O’Connor (1940) stated that sexava outbreaks were always worst and most frequently occurred in areas, which experienced no regular dry season during the year. Room et al. (1984) monitored a population of S. uniformis on coconuts over a five-year period. They constructed static life tables and calculated k values for the duration of the study. The analysis suggested that S. uniformis populations were sensitive to mortality in the crown of the palm. This mortality appeared to be related to the duration of periods without rainfall. The number of first instar nymphs ascending the palm trunks was positively correlated with the largest number of consecutive days in a month without rain, that is more eggs hatched when rain followed a dry spell. Young (1984, 1985 and 1987a), working with S. decoratus on coconuts, described two types of outbreak (Figure 1). Firstly, localised outbreaks in areas with an evenly distributed rainfall throughout the year. This type of outbreak commonly occurs on overgrown coconut blocks and is typified by a slow population increase over 12 to 18 months, e.g. Bubia near Lae. The generations are not discrete. Secondly, a wide spread outbreak preceded by a drought or a prolonged dry spell. The outbreak is typified by a synchronised egg hatch and rapid population increase after normal rainfall resumes, for example, in the Bogia District of the Madang province, where prolonged dry spells often occur between May and October. Initial egg diapause leading to discrete generations explains the second type of outbreak (Young 1990). 35 Low population of S. decoratus e.g. less than 4 adults/mature palm Coconut block becomes overgrown. Grass and weeds up to 1.5 metres under palms Dry spell. Females continue to lay eggs which owing to low soil moisture go into an initial diapause Palms recover High humidity beneath undergrowth creates conditions favourable to survival of eggs and first instar nymphs Rain at end of dry spell. Eggs start to take up water. The development of eggs is synchronised Populations of S. decoratus decline Population of S. decoratus increase over 3 generations with resultant defoliation of palms 7 - 8 weeks later eggs hatch producing a discrete generation of first instar nymphs. The nymphs climb the palm and start to feed. Improved plantation management and/or influence of natural enemies e.g. S. dallatorreanum Severe defoliation of palms 4 - 5 months after egg hatch S. decoratus exhaust food supply Defoliation of palms becomes so severe that the palms are unable to nourish coconuts. Coconuts fall before maturity Generations remain discrete for up to 3 generations Fig. 1. A suggested dioristic model of S. decoratus outbreaks on coconut palms. 36 Natural Enemies Predators The following vertebrates have been recorded feeding on sexava nymphs and adults: the Torresian crow, Corvus orru Bonaparte; Brahminy kite, Haliastur indus (Boddaert); willie wagtail, Rhipidura leucophrys (Latham); goshawk, Accipter sp.; crested hawk, Aviceda subceristata (Gould); sacred kingfisher, Halcyon sanctus Vig. and Horsf.; two lizards, the green skink, Lamprolepis smaragdina (Barbour) and the gecko, Gehydra oceanica (Lesson); the terrestrial frog, Platymantis papuensis Meyer (G. Baloch pers. comm.; S. Embupa, pers. comm.; Froggatt, 1935; Room et al. 1984; Young, unpublished data). Froggatt (1935) considered that vertebrate predators were not numerous enough to influence populations of sexava - a perception that is probably correct, since it has not been seriously challenged by entomologists since that time. The predatory ant, Oecophylla smaragdina (Fabricius), was thought to be a potential predator of sexava and unsuccessful attempts were made to introduce the ant into some islands of the Manus province where it does not occur naturally (P. Room pers. comm.). Young (1984) found that O. smargdina was not a significant predator of sexava. Parasites Egg parasites The eggs of sexava are parasitised by a range of hymenopteran parasites (Table, 3). In 1933 Froggatt travelled to Ambon in Eastern Indonesia and returned to the Manus research station with a colony of the egg parasite, Leefmansia bicolor Waterston (Hymenoptera: Encyrtidae). The L. bicolor colony was multiplied at the research station and released on Manus and New Hanover, where it became established. Another species of encyrtid was recovered from sexava eggs on New Hanover (Froggatt, 1935). Froggatt (1935) stated the parasite was very similar to L. bicolor, but as far as is known the status of the native New Hanover parasite has not been determined. Froggatt (1937) considered only three species of egg parasite were of economic importance namely; L. bicolor, Doirania leefmansi Waterston (Hymenoptera: Trichogrammatidae) and an unidentified mymarid from New Hanover. 37 Table 3. Parasites recorded from sexava eggs FAMILY SPECIES DISTRIBUTION REFERENCE Encyrtidae Leefmansia bicolor Waterston Manus Island, New Hanover Froggatt (1935); (1937); (1938) Froggatt and O’Connor (1940) O’Connor (1937) New Hanover Froggatt (1935) Tetrastichus dubius (Waterston) Manus Island, New Hanover, New Ireland and New Britain (Islands of the Bismarck Archipelago). Froggatt (1937) Froggatt and O’Connor (1940) Tetrastichus nr dubius Bogia coast of the Madang Province Young (1987a) Tetrastichus sp Bubia, Morobe Province Young (1987a) Mymaridae Anneckia oophaga Rao Platypatasson fransseni Ogloblin Anaphes sp Stethynium sp. Islands of the Bismarch Archipelago Embupa (pers. comm.) Trichogrammatidae Doirania leefmansi Waterston New Hanover Froggatt (1935); (1937); (1938) Froggatt and O’Connor (1940) O’Connor (1937) Scelionidae Triteleia atrella (Dodd) Islands of the Bismarch Archipelago and mainland PNG Froggatt (1935); (1937) Froggatt and O’Connor (1940) Young (1987a) ? Eulophidae Parasites of nymphs and adults Two parasites have been recorded from sexava nymphs and adults: Stichotrema dallatorreanum Hofender (Strepsiptera: Myrmecolacidae) and Exorista notabilis (Wilkinson) (Diptera: Tachinidae) (O’Connor 1959; Young 1987a and 1987b). E. notabilis was not considered an efficient parasite at Bubia since the number of adult S. decoratus parasitised by the fly did not exceed ten percent (unpublished data). The most important parasite is S. dallatorreanum (Young 1984). S. dallatorreanum is recorded from mainland PNG. It has a complex life history with females parasitising three species of sexava and the males parasitising ants. The sexava attacked are Sexava nubila, Segestes decoratus and Segestidia novaeguineae (J. Ardley pers comm.; O’Connor 1959; Young 1987b). The ant host is probably Camponotus papua Emery, which is more correctly known as Camponotus novaehollandiae group Mayr (A. Andersen pers. comm.; R. Taylor pers. comm.; 38 Young 1987b). The male S. dallatorreanum was originally described from a specimen captured at light, as Caenocholax (Rhipidocolax) acutipennis by Kogan and Oliveira (1964). Y. Hirashima (pers. comm.) examined a male captured en copula with a female S. dallatorreanum, he confirmed the view of Luna de Carvalho (1972) that C. acutipennis was the male of S. dallatorreanum (Young 1987b). The cephalothorax of the mature female S. dallatorreanum is extruded through the abdominal wall of the host (Young 1987a). A free-living adult male flies to the parasitised host and settles on the extruded parasite’s cephalothorax. The male mates by inserting his aedeagus through the operculum of the brood canal, which is part of the female’s cephalothorax. Three to five weeks later the operculum ruptures and first instar or primary larvae emerge from the brood canal (Kathirithamby 1989; Luna de Carvalho and Kogan 1991). The primary larvae show a distinct sexual dimorphism, which was first described by Carvalho (1959) and later confirmed by Young (1987b). The female primary larva has a bilobed labrum, a large buccal orifice and a labial plate half as long as it is wide. Eyes large, situated dorsolaterally and midway between the tip of the head and the thorax (Figure 2). There are very short setae on the dorsal surface of the ninth abdominal segment. The male primary larva is one third smaller than the female, labrum continuous and a small buccal orifice. The flat labial plate is of equal length and breadth, the nineth abdominal segment has two setae on the dorsal surface extending beyond the base of the large caudal styles. Eyes, smaller than the female, are situated dorso-laterally, well back on the head close to the thorax (Figure 3). Figure 2. Head and thorax of female primary larva showing bilobed labrum and large eye situated dorso-ventrally midway between the tip of the head and the thorax ≅ X 260 Figure 3. Head and thorax of male primary larva showing continuous labrum and small eye situated dorso-ventrally close to the thorax ≅ x 200 The male primary larva is believed to enter the ant larva and undergo its immature stages in the larva and pupa before emerging as a winged adult from the abdomen of the adult ant (Carvalho 1959; Young 1987b). The female primary larva penetrates the abdominal cuticle of a sexava host, undergoes ecdysis and the legless larva continues to grow inside the host's abdominal cavity. Mean time from penetration by the primary larva until extrusion of the adult female's cephalothorax from the host is 53.5 days (Young 1987b). Young (1987a) found that first instar S. decoratus nymphs infected with primary larvae were dead within 45 days. Parasitised adult female hosts matured 67 percent fewer eggs than unparasitised hosts. Parasitism drastically reduced the fitness and therefore longevity of the adult host (O’Connor 1959; J. Ardley pers. comm.). It would appear that S. dallatorreanum 39 controls mainland sexava populations by reducing fecundity, both by direct reduction of egg production and reducing the time over which eggs can be laid. There are no records of either nematode or microsporidian parasites from sexava. Control Methods Mechanical control Froggatt and O’Connor (1940) reported that the smoke from fire lit under palms caused adults and nymphs to fall down to the ground, where they were collected and destroyed - a method which is still occasionally used in village coconut groves. Smoke will not dislodge sexava from tall palms and shorter palms can be damaged by fire. Froggatt (1935) applied sticky bands of “Tanglefoot” around coconut palm trunks. He found the bands trapped many first and second instar nymphs. He concluded that this method might have economic possibilities in small areas at the start of an outbreak, but would be too expensive on a large scale (Froggatt 1938). More recently the polybutene based products, Osticon® (also containing 15% trichlorfon) and Tacktrap® have been used on coconut trunks as sticky bands with much the same results as those of Froggatt (Young unpublished data). Bands of polybutene have the disadvantage of causing shrinkage of the trunk where the sticky bands cover the bark and splitting of the bark adjacent to the bands (Young 1996). Adult sexava have no trouble jumping over sticky bands when descending from the palm (Young 1984). Both Osticon® and Tacktrap® are expensive and for the results achieved, uneconomic (unpublished data). Raking and hoeing of soil under the palms in order to expose eggs to parasites and desiccation has long been advocated (Froggatt 1935). There is no evidence that this practice has any influence on sexava populations (unpublished data). Chemical control Froggatt (1935) reported that paris green and bran baits were not attractive to sexava. Additionally, he mixed either sodium arsenite or mercuric chloride or sodium fluoride or paris green with “Tanglefoot”, smeared the mixture on boards and allowed sexava nymphs and adults to walk across the boards. Each treatment gave 100 percent mortality, but the cost of application in the field was relatively high (Froggatt 1935). Laboratory trials with 5% arsenical dusts gave over 80% mortality, while the same concentration of derris and pyrethrum gave poor and variable results (Froggatt 1938). Froggatt and O’Connor (1940) stated that while arsenical dusts caused high mortality to sexava in the laboratory, application in the field was impractical. In 1958, M. Catley (pers. comm.) reported good control of sexava on New Hanover using 10 percent BHC dust applied through a power driven dusting machine. G. Baloch (pers. comm.), C. Perry (pers. comm.) and Young (1981) have suggested spraying the base of the palm trunk with chlordane or dieldrin to prevent first instar nymphs ascending the palm. Young (1984) later pointed out it was better to spray the soil surface around the base of the trunk. The insecticide remained active in the soil for up to three months as against three to five days on the trunk. It is doubtful if the method was economic, in addition to which both insecticides have since been banned. Aerial spraying with malathion ULV to control sexava has had variable success. T. Bourke (pers. comm.) reported that while good mortality of adults was achieved after aerial spraying coconuts in the Manus province during 1970, populations returned to high levels six to eight weeks later. C. Perry (pers. comm.) reported good control when Asuramba plantation was aerially sprayed with malathion ULV in August 1974. There was concern that aerially spraying with malathion ULV would affect the parasite/predator complex in coconut palms (C. Perry pers. comm.). Trunk injection involves boring a 15 cm x 1.5 cm hole at an angle of 45 degrees into the palm trunk. The hole is injected with 10 g a.i. of monocrotophos e.c. and is then plugged with a dowel. Since monocrotophos is systemic the chemical is translocated into the leaflets where it poisons feeding sexava. Monocrotophos remains at insecticidal levels in the leaflets for about 40 12 weeks. A second injection is carried out at 12 weeks to control those sexava nymphs, which have hatched after the initial 12 weeks. Trunk injection is the only economic means of chemical control that has been used on oil palm, where it has been in use for the past 20 years (S. Embupa pers. comm.; C. Perry pers. comm.; R. Prior pers. comm.). However, the low price of copra probably makes the use of trunk injection on coconuts uneconomic. Biological control Most cultural, chemical and mechanical methods proved to be either impractical or uneconomic. For these reasons much research over the past 55 years has been devoted to biological control (Young, 1987a). Initially, egg parasites received the most attention (Froggatt 1935, 1937, 1938; Froggatt and O’Connor 1940). Consignments of sexava eggs parasitised by L. bicolor, D. leefmansi and Stethynium sp. were sent from New Hanover to most coconut growing areas of the Northern New Guinea mainland and the islands of the Bismarck Archipelago (Froggatt 1935, 1937, 1938; Froggatt and O’Connor 1940). O’Connor (1937) stated that, in New Hanover, egg parasites appeared to keep sexava infestations at a low level. After World War II, regular releases of L. bicolor and D. leefmansi were made on coconut plantations in numerous localities and more recently on oil palm in West New Britain (J. Ardley, pers. comm.; S. Embupa pers. comm.). J. Ardley (pers. comm.) doubted that egg parasites could be relied on as a means of biological control. Despite the claims that egg parasites controlled sexava populations in New Hanover, A. Catley (pers. comm.) recorded sexava outbreaks on coconuts in New Hanover during the 1950s. L. bicolor and D. leefmansi both show higher rates of parasitism of eggs laid in epiphytes and the coconut crown than eggs laid in soil (Franssen 1954; Froggatt 1935; Froggatt and O’Connor 1940). Young (1987a) found that T. atrella and Tetrastichus spp. could not locate eggs buried in soil. Since the majority of eggs are laid in the soil, there is a need to establish if L. bicolor and D. leefmansi can parasitise eggs laid in soil. Room et al. (1984) found that egg mortality had little effect on S. uniformis populations in Manus. O’Connor (1959) first suggested the potential of parasites of sexava nymphs and adults. In 1941 he attempted to introduce S. dallatorreanum to Pak island, Manus province, by releasing several hundred parasitised sexava which, he had collected on the mainland. When O’Connor later visited Pak Island in 1945 he found no sign of the parasite (O’Connor 1959). Before O’Connor left Manus in 1941 he collected nymphs and adult S. uniformis and placed them in cages, which had been previously occupied by the parasitised mainland sexava prior to releasing the hosts on Pak Island. The S. uniformis apparently became infected with primary larvae left in the cages. The parasitised hosts were returned to Rabaul where the cephalothoraxes of the female parasites were extruded through the body walls of the hosts. The hosts died without any of the female parasites producing primary larvae (O’Connor 1959). Later, J. Ardley (pers. comm.) sent parasitised sexava from Bainyik, East Sepik, to Keravat, East New Britain, where G. Dunn raised the female parasites in S. defoliaria defoliaria hosts. Once again the cephalothoraxes of the female parasites were extruded from the host but no primary larvae were produced. Ardley (pers. comm.) remarked that both attempts to introduce the parasite into off shore islands were predicated on the belief that S. dallatorreanum was parthenogenetic. He concluded that the failure of female parasites to produce primary larvae was a result of their eggs not being fertilised by males. S. dallatorreanum has been recorded parasitising S. decoratus, S. novaeguineae and S. nubila from mainland PNG. There appears to be good prospects of introducing the parasite to island populations of S. decoratus, S. defoliaria defoliaria and S. uniformis, although there are two aspects of the biology of host and parasite that need further investigation. Firstly, more work is required on the biology of the male parasite and its ant host, C. novaehollandiae. Surveys of the distribution and abundance of C. novaehollandiae in the target localities would be necessary to determine if S. dallatorreanum had any chance of survival in areas where it does not occur naturally. Secondly, the role of egg diapause in the population dynamics of the target host needs further study. When the host population increases rapidly in one generation it could take S. dallatorreanum two or three generations to exert control over the host population. Even if the parasite became established it might not give the level of control found on mainland PNG. 41 Acknowledgements Thanks are due to Dr. R. Peng, post doctoral fellow at the Northern Territory University, Darwin, for his helpful comments on the manuscript and Mrs C. Gilkeson EM Unit, School of Biological Sciences, Sydney Technical College for Figures 2 and 3. References Bailey, P., O’Sullivan, D. and Perry, C. (1977). Effect of artificial defoliation on coconut yields in Papua New Guinea. Papua New Guin. A J. 28 (2, 3 and 4), 39-45. Bourke, R. M. (1996). Edible indigenous nuts in Papua New Guinea. In Stevens, M. L., Bourke, R. M. and Evans, B. R. Eds. South Pacific Indigenous Nuts. ACIAR proceedings No. 69. Canberra, ACT. 45-55. Franssen, C. J. H. (1954). Biologische bestrijding van de sabelsprinkhaan Sexava nublia St. op de Talaude-eilanden. Ent. Ber. Amst. 15, 99-102. Froggatt, J.L. (1935). The long horned tree hopper of coconut Sexava spp. New Guinea agric. Gaz. 1, 16-17. Froggatt, J.L. (1937). Egg parasites of Sexava spp. in the Territory of New Guinea. New Guinea Agric. Gaz. 3 (2), 24-25. Froggatt, J. L. (1938). Measures for the control of the coconut tree hopper (Sexava spp.). New Guinea Agric. Gaz. 4 (3), 3-6. Froggatt, J.L. and O’Connor, B. A. (1940). Insects associated with the coconut palm. New Guinea Agric. Gaz. 6 (3), 16-32. Hartley, J. C. (1990). Egg biology of the Tettigoniidae. In Bailey, W. J. and Rentz, D. C. F. Eds. The Tettigoniidae: Biology, Systematics and Evolution. Crawford house press, Bathurst, Australia, pp. 41-70. Kathirithamby, J. (1989). Review of the order Strepsiptera. Systematic Entomology 14, 41-92. Kogan, M. and Oliveira, S. J. (1964). New Guinean Mengeidae and Myrmecolacidae of the American Museum of Natural History (Strepsiptera). Studia Ent. 7: 459-470. Luna de Carvalho, E. (1959). Segunda contribuição para o estudo dos estrepsípteros angolenses (Insecta: Strepsiptera). Publções cult. Co. Diam. Angola no. 41, 125-154. Luna de Carvalho, E. (1972). Algumas considerações sobre Mirmecolacídeos da Nova Guiné (Insecta Strepsiptera). Ciência Biológica, (Portugal) 1: 1-6. Luna de Carvalho, E. and Kogan, M. (1991). Order Strepsiptera, in Stehr, F. W. Ed. Immature Insects Vol.2. Kendal/Hunt Pub. Co. Dubuque, Iowa, pp 659-673. O’Connor, B. A. (1937). Progress of work on Sexava spp., the coconut treehopper. New Guinea Agric. Gaz. 3 (1), 1-4. O’Connor, B. A. (1959). The coconut treehopper, Sexava spp., and its parasites in the Madang district. Papua New Guin. Agric. J. 11 (4), 121-125. Rentz, D. (1996). Grasshopper country, the abundant orthopteroid insects of Australia. University of NSW Press, Sydney, Australia. P 105. 42 Room, P. M., Perry, C. H. and Bailey P.T. (1984). A population study of the coconut pest Segestidea uniformis (Willemse) (Orthoptera: Tettigoniidae) on an equatorial island. Bull. ent. Res. 74, 439-451. Willemse, F. (1977). Classification and distribution of the Sexavae of the Melanesian subregion (Orthoptera, Tettigonioidae, Mecopodinae). Tijdschr. Ent. 120, 213-277. Willemse, F. (1979). Additional notes on the Sexavae of the Melanesian subregion (Orthoptera, Tettigonioidae, Mecopodinae). Ent. Ber., Amst. 39, 4-9. Young, G. R. (1981). Sexava: a pest of coconut and oil palm. 3 pp. Lae, Papua New Guinea, Dept. of Primary Ind. (Bubia, information bulletin no. 37). Young, G. R. (1984). Sexava outbreaks on coconuts in northeast mainland PNG and some islands of the Bismarck Archipelago. 11 pp. Lae, Papua New Guinea, Dept. Primary Ind. (Bubia, information bulletin no. 48). Young, G. R. (1985). Observations on the biology of Segestes decoratus Redtenbacher (Orthoptera: Tettigoniidae), a pest of coconut in Papua New Guinea. Gen. Appl. Ent. 17, 57-64. Young, G. R. (1987a). Some parasites of Segestes decoratus Redtenbacher (Orthoptera: Tettigoniidae) and their possible use in the biological control of tettigoniid pests of coconuts in Papua New Guinea. Bull. Ent. Res. 77, 515-524. Young, G. R. (1987b). Notes on the life history of Stichotrema dallatorreanum Hofender (Strepsiptera: Myrmecolacidae) a parasite of Segestes decoratus Redtenbacher (Orthoptera: Tettigoniidae) from Papua New Guinea. Gen. Appl. Ent. 19, 57-64. Young, G. R. (1990). Water uptake by the eggs of Segestes decoratus Redtenbacher (Orthoptera: Tettigoniidae: Mecopodinae). Gen. Appl. Ent. 22, 17-19. Young, G. R. (1996). An association between the crazy ant Anoplolepis longipes (Jerdon) (Hymenoptera: Formicidae) and the coconut spathe moth Tirathaba rufivena (Walker) (Lepidoptera: Pyralidae) on coconut palms in the Morobe Province of Papua New Guinea. 2. The effect on yield and nut shedding of ant and moth exclusion. Papua New Guinea J. Agric. For. Fish. 39 (1), 7-11. 43 INTEGRATION OF CHEMICAL, CULTURAL AND BIOLOGICAL CONTROLS IN NORTH QUEENSLAND SUGARCANE FOR GREYBACK CANEGRUB, DERMOLEPIDA ALBOHIRTUM (WATERHOUSE) (COLEOPTERA: SCARABAEIDAE) Les Robertson Bureau of Sugar Experiment Stations PO Box 566 Tully Queensland 4854 Abstract The soil-living pest of sugarcane, greyback canegrub, has been managed for 50 years with insecticides. Following withdrawal of the organochlorine insecticides, populations of the pest have increased. Control of greyback canegrub requires combined use of insecticide, cultural practices and biological controls. Cultural practices which reduce damage in individual fields include manipulation of planting and harvesting dates, use of trap crops to attract ovipositing beetles, tolerant varieties, retention of all crop residues as a trash blanket on the soil surface, control of grass weeds, and avoiding applying lime close to the time of application of the registered chlorpyifos formulation. Biological control may be enhanced by minimum tillage planting, no-till ratooning, extended crop cycles, and augmentation with the biological insecticide based on Metarhizium anisopliae. A decision-aid diagram is presented which allows assessment of risk of greyback canegrub damage for farms and individual canefields, as well as gives options to reduce the impact of greyback canegrub damage. Key words: sugar cane, greyback cane grub, cultural practices, biological control Introduction Greyback canegrub, Dermolepida albohirtum (Scarabaeidae: Melolonthinae), is the most damaging of the insect pests of sugarcane in Australia. It is endemic to northeastern Queensland and southern Papua New Guinea, and occurs in sugarcane from Sarina in central Queensland to Mossman in far north Queensland. Infestations were also detected recently in the new sugar-growing area around Mareeba on the Atherton Tableland. The pest usually occurs on free-draining alluvial and volcanic soils. Greyback canegrub is univoltine and beetles emerge with the onset of early summer rains, or with irrigation from October each year. The beetles fly at dawn and dusk, seeking trees on which to feed, and tall cane and grasses under which to lay eggs. The larvae feed on the roots of sugarcane and other grasses generally from January until July. Grub damage to sugarcane roots causes reduced growth and yield of crops, dislodgement of the root system from the soil, poor regrowth of ratoon crops after harvest, and premature replacement of the crop. Only one insecticide, controlled-release chlorpyrifos, is available for the control of greyback canegrubs. This material is applied to the soil in the plant crop, and is registered for control of greyback canegrub for only one year. Subsequent ratoon crops may be unprotected. The average crop cycle for sugarcane in Queensland is five years (plant crop and four ratoons). Widespread outbreaks of greyback canegrub have occurred in sugarcane throughout the range of the pest since 1990, with over 3,000 ha damaged each year since 1996. The outbreaks followed withdrawal of organochlorine insecticides in 1986. Benzene hexachloride (BHC) had been used since 1947 to prevent increases of grubs in sugarcane, and 26,000 ha of sugarcane were treated annually with BHC (Wilson 1969). Heptachlor was also used as a preventative treatment to control greyback canegrub on up to 3,000 ha per year, predominantly in the Burdekin River district. Resurgence of pests is common when pesticide use is stopped or becomes less effective. 45 During the 40 years when greyback canegrub was under good control with insecticides, there was little research on alternative biological and cultural controls. The impetus for finding alternatives to insecticides came when the controlled-release chlorpyrifos product became ineffective on many farms in the Burdekin River district. This was a result of enhanced microbial degradation in the alkaline soils of that area (Robertson et al. 1998). Organophosphate and carbamate insecticides generally are susceptible to enhanced microbial degradation when used repeatedly (Felsot 1989), and alkaline conditions also predispose organophosphates to accelerated breakdown (Racke et al. 1990). Intensive research on non-chemical controls for greyback canegrub was initiated by BSES in the early 1990s, and this paper is a review of progress towards sustainable control of the pest. Effect of Farming Practices on Greyback Canegrub Populations Cane height Greyback canegrub beetles preferentially lay their eggs in soil beneath tall crops of sugarcane rather than under short cane (Illingworth 1918). Cane on ridges may also attract more ovipositing beetles compared to similar-sized cane on lower parts of the same block. Cane planted early in the season and crops harvested early, tend to be taller than crops planted or harvested late. The effect of this on grub infestation was studied by Ward and Cook (1996) in the Burdekin River district. They showed that most damage was suffered in early-planted cane and in ratoons harvested early in the season. This research has led to widespread adoption of late planting in the Burdekin, and subsequent concentration of damage on less valuable, earlyharvested, older ratoons. Beetle preference for tall crops can be used to attract a high proportion of ovipositing females to “trap crop” areas of early-planted or early-harvested strips of cane, or to tall-growing cultivars of cane within blocks of a shorter cultivar. Cocco and Robertson (unpublished) recorded three times as many grubs in the rapidly growing cultivar Q127 compared to Q96 or Q117 in the same block (see also Robertson and Cocco 1998). Similarly, higher densities of grubs were found in cane strips harvested early for planting material and then ratooned, compared to the bulk of the blocks harvested later in the season. Future research aims to identify the proportion or area of a field which needs to be managed as a trap crop, to reduce damage on the remainder of the field. Manipulation of planting and harvesting dates and the use of trap crops may not reduce the population density of greyback canegrub at the farm or district level, unless trap crops are destroyed by cultivation when grubs are actively feeding in the root zone. These strategies minimise economic losses by reducing damage on the most valuable crops. Trash blanketing and no-till ratooning Cane is usually managed in one of two ways. Cane can be burnt before harvest to remove excess leaf material and remaining trash may also be burnt; weeds in the subsequent ratoon crop may be controlled by cultivating the inter-row space. The second method involves harvesting the green crop without prior burning, and retaining all of the crop residues as a blanket on the soil surface. Weeds tend to be smothered by the trash blanket, but some weed control may be required using herbicides. Robertson and Walker (1996) found that fewer greyback canegrubs survived under green cane trash blankets compared to burnt and cultivated ratoons. The mechanism leading to more grubs under burnt conditions is not known, although grubs grew more slowly under green cane trash blankets, possibly because of cooler soil conditions compared to that under bare soil surfaces (Robertson and Walker 1996). Beetles oviposit equally under trash blankets and in burnt and cultivated ratoons. Control of grass weeds Greyback canegrub beetles are attracted to tall-growing grasses as well as sugarcane, and sugarcane fields with abundant grass weeds may become severely infested irrespective of trash blanketing (Robertson and Walker 1996) or late harvesting (unpublished observations). Guinea grass (Panicum maximum) is common in Queensland cane growing areas and, if allowed to infest canefields, can contribute to greyback canegrub infestation by attracting ovipositing beetles. 46 Tolerant cultivars Some cultivars of cane are able to tolerate damage by greyback canegrub better than others. In the wet tropics, Q158 showed least damage, Q138 intermediate damage and Q152 severe damage when infested with similar numbers of greyback canegrubs in the same block (Anon. 1996). The relative size of the root system is inversely correlated with the level of damage inflicted by canegrubs (P.G. Allsopp, unpublished data). Cane varieties with resistance to canegrubs are being sought (Allsopp et al. 1997). Effect of farming practices on insecticidal control Chlorpyrifos is rapidly degraded under alkaline conditions (Racke et al. 1990). Application of lime or other alkaline amendments and fertilisers to soil about the time when the canegrub insecticide is also applied may result in accelerated loss of active ingredient from the controlledrelease formulation of chlorpyrifos (Chandler 1997). Lime should be applied well before replanting (when insecticide is applied) to improve the efficacy of the insecticide. Biological Control Fungal biological insecticide The green muscardine fungus (Metarhizium anisopliae) has been known to infect greyback canegrubs since early this century (Illingworth and Dodd 1921). A virulent strain of this pathogen was collected from greyback canegrub at Tully, cultured on rice, and tested as a biological insecticide (Robertson et al. 1997a). Application of 33 kg of product per ha gave 50 to 60% control of greyback canegrubs in the plant crop before the period when damage to cane becomes evident (Robertson et al. 1997a). A single application in plant cane also controlled 70% of the grubs in each of the second and third years, suggesting that the material persists or is propagated when grubs die of the disease. A commercial product (BioCaneTM, Bio-Care Technology Pty Ltd) is currently being evaluated extensively under an experimental trial permit in north Queensland sugarcane. Incidence of entomopathogens A survey and monitoring program of population changes in greyback canegrub has been conducted since 1993 (Robertson et al. 1997b). Several microorganisms were identified in greyback canegrubs, which died after collection during the monitoring program (Dall et al. 1995). A coccidian protozoan, Adelina sp., was prevalent in greyback canegrubs at sites where grub numbers declined within and between seasons (Robertson et al. 1997b). The incidence of this pathogen was highly variable, with grubs from the Burdekin River district consistently showing 1% or lower incidence, while grubs collected from the Herbert River district 160 km to the north had up to 88% infection by Adelina sp. Metarhizium incidence varied less than Adelina, with a relatively constant 30% mortality each year at a site near Tully despite changes in population density of the host. Other diseases including Bacillus popilliae and a microsporidian near Nosema sp. were rare (Dall et al. 1995). Crashes in populations of greyback canegrub were recorded at several locations and grubs at all these locations had a relatively high incidence of Adelina infection (Robertson et al. 1997b). Conversely, the persistent outbreak in the Burdekin River district is characterised by grubs with very low levels of entomopathogens. Prior to the widespread use of organochlorines, cycles of severe damage by greyback canegrub, followed by years of little or no damage, were recorded in annual reports of the sugar industry (Robertson et al. 1997b). It is now thought that entomopathogens such as Adelina were responsible for crashes in greyback populations, and reduction in damage, in the pre-insecticide era. Effect of farming practices on incidence of entomopathogens Frequent cultivation may suppress the diseases of grubs by dispersing the infective spores or otherwise reducing their viability. The pathogen Adelina is prevalent in greyback canegrubs in ratoons, which are managed by green cane trash blanketing and no-till ratooning, and is less common in burnt and cultivated ratoons. It is not common in plant cane, which is established after cultivation. Cane in the Burdekin River district is almost invariably burnt before harvest, ratoons are cultivated and crop cycles are generally short. Incidence of all pathogens was low in grubs collected from the Burdekin over the past four years and grub numbers have remained 47 high since 1990. Cultivation was believed responsible for reducing the incidence of the pathogens in the melolonthine scarab Costelytra zealandica in pastures (Miln 1982). Minimum tillage planting is being evaluated for its ability to maintain effective levels of entomopathogens for control of greyback canegrub in the plant crop. The old cane crop is killed with herbicide and fallowed without cultivation until the next crop is established. Samson and Phillips (1997) showed that minimum tillage following a herbicide fallow resulted in higher densities of Metarhizium anisopliae spores remaining in the soil after replanting compared to conventionally cultivated soil. Integrated Pest Management Integration of chemical, cultural and biological controls is necessary to achieve control of greyback canegrub in the absence of the persistent organochlorine insecticides. Except in the Burdekin River district, controlled-release chlorpyrifos should be applied during outbreaks, to reduce damage in the plant crop. Care should be taken to apply lime products several months before applying the insecticide. In the Burdekin where the insecticide has failed to protect crops, early planting should be avoided. Older, less valuable crops should be harvested early and ratooned to attract ovipositing beetles away from the more productive young crops. Trap crops may assist in reducing overall damage to cane on a farm. Cultivation should be kept to a minimum when replacing crops, to preserve entomopathogens from one crop cycle to the next. Where possible, herbicide should be used to kill the old crop, rather than cultivation. Once the old root system has decomposed, the soil is generally friable enough along the row to allow replanting after one or two cultivations. Plant pathogens may also persist with minimum tillage, and resistant cultivars need to be planted to avoid possible losses to plant diseases. Tolerant cultivars such as Q158, where available, should be planted when damage has been severe. Varieties with a small root mass, such as Q152 in the wet tropics, and Eos in the Burdekin, should be avoided in grub-prone areas. All crops should be cut green and trash retained as an undisturbed blanket on the soil surface. In addition, crops should be ratooned for as long as possible and profitable. These strategies apparently favour the increase in entomopathogens in greyback canegrub, leading to natural control of the pest. When commercially available, the biological insecticide based on Metarhizium should be applied to give at least partial protection of plant and ratoon crops. This product will be especially useful in the Burdekin where chlorpyrifos is commonly ineffective, and where levels of Metarhizium are presently low. A decision-aid diagram to assess the risk of infestation by greyback canegrub at the farm and canefield level, and to give options to reduce infestations and damage, is given in Figure 1. Adoption of IPM for control of greyback canegrub has been slow, although components of the strategies have been implemented by some cane growers (e.g. late planting in the Burdekin). The framework presented here is being promoted throughout north Queensland to develop and extend IPM for greyback canegrub (e.g. see Robertson and Cocco 1998). 48 History of greyback canegrub damage yes Green cane trash blanketed ratoons yes no Recently developed loamy soil yes no no Intensive cultivation, in fallow or before planting yes no HIGH RISK Chlorpyrifos granules applied in plant crop no yes Lime applied at planting, or high soil pH yes no Late planted, or late harvested ratoon yes Abundant grass weeds in crop no Grub-tolerant variety of cane adopt chemical cultural and biological controls for greyback canegrub no yes no LOW RISK Figure 1. Decision-aid diagram to assess risk of damaging infestations of greyback canegrub in individual canefields in north Queensland Acknowledgements The Sugar Research and Development Corporation funded much of the research reported in this paper. Numerous canegrowers in north Queensland have cooperated with on-farm trials and evaluation of pest management strategies. This paper is presented with the permission of the Director, Bureau of Sugar Experiment Stations. References Allsopp, P.G., Shepherd, K.M., Smith, G.R., Cox, M.C. and Robertson, S.K. (1997). Use of hostplant resistance as a component in IPM for canegrubs. Pp. 102-5 in Allsopp, P.G., Rogers, D.J. and Robertson, L.N. (eds.), ‘Soil Invertebrates in 1997’. Bureau of Sugar Experiment Stations, Brisbane. Anon. (1996). BSES Bull. 56: 24. Chandler, K.J. (1997). Alkaline soil amendments jeopardise soil-insect control with chlorpyrifos. Pp.117-8 in Allsopp, P.G., Rogers, D.J. and Robertson, L.N. (eds.), ‘Soil Invertebrates in 1997’. Bureau of Sugar Experiment Stations, Brisbane. Dall, D., Lai-Fook, J., Robertson, L. and Walker, P. (1995). Microorganisms associated with mortality of greyback canegrubs. Proc. Aust. Soc. Sugar Cane Technol. 17: 106-9. Felsot, A.S. (1989). Enhanced biodegradation of insecticides in soil: implications for agroecosystems. Ann. Rev. Entomol. 34: 453-76. Illingworth, J.F. (1918). Cane grub investigation. Div. Entomol. Bull., BSES 8: 5-7. 49 Illingworth, J.F. and Dodd, A.P. (1921). Australian sugar cane beetles and their allies. Div. Entomol. Bull., BSES 16: 1-104. Miln, A.J. (1982). The effects of cultivation on disease incidence in grass grub populations. Proc. NZ Weed and Pest Control Conf. 35: 86-9. Racke, K.D., Laskowski, D.A. and Schultz, M.R. (1990). Resistance of chlorpyrifos to enhanced biodegradation in soil. J. Agric. Food Chem. 38: 1430-6. Robertson, L.N., Chandler, K.J., Stickley, B.D.A., Cocco, R.F. and Ahmetagic, M. (1998). Enhanced microbial degradation implicated in rapid loss of chlorpyrifos from the controlledrelease formulation suSCon Blue in soil. Crop Prot. 16: (in press). Robertson, L. and Cocco, R. (1998). Greyback canegrub - why do we still have a problem? BSES Info. Meet. Burdekin, Feb. 1998, Bureau of Sugar Experiment Stations, 9-12. Robertson, L.N., Dall, D.J., Lai-Fook, J. and Kettle, C.G. (1997b). Population dynamics of greyback canegrub (Dermolepida albohirtum) in north Queensland sugarcane (Scarabaeidae: Melolonthinae). Pp. 140-4 in Allsopp, P.G., Rogers, D.J. and Robertson, L.N. (eds.), ‘Soil Invertebrates in 1997’. Bureau of Sugar Experiment Stations, Brisbane. Robertson, L.N., Kettle, C.G. and Bakker, P. (1997a). Field evaluation of Metarhizium anisopliae for control of greyback canegrub (Dermolepida albohirtum) in north Queensland sugarcane. Proc. Aust. Soc. Sugar Cane Technol. 19: 111-7. Robertson, L.N. and Walker, P.W. (1996). Effect of green-cane harvesting and trash blanketing on numbers of greyback canegrub. Proc. Aust. Soc. Sugar Cane Technol. 18: 78-81. Samson, P.R. and Phillips, L.M. (1997). Farming practices to manage populations of sugarcane soldier fly, Inopus rubriceps (Macquart), in sugarcane. Pp. 96-101 in Allsopp, P.G., Rogers, D.J. and Robertson, L.N. (eds.), ‘Soil Invertebrates in 1997’. Bureau of Sugar Experiment Stations, Brisbane. Ward, A.L. and Cook, I.M. (1996). Effect of planting and harvesting date on greyback canegrub damage in the Burdekin River area. Pp. 226-7 in Wilson, J.R., Hogarth, D.M., Campbell, J.A. and Garside, A.L. (eds.). ‘Sugarcane: Research Towards Efficient and Sustainable Production’. CSIRO Tropical Crops and Pastures, Brisbane. Wilson, G. (1969). Insecticides for the control of soil-inhabiting pests of sugar cane. Pp. 259-82 in Williams, J.R., Metcalf, J.R., Mungomery, R.W. and Mathes, R. (eds.). ‘Pests of Sugar Cane’. Elsevier, Amsterdam. 50 BAITS FOR FRUITPIERCING MOTHS - THE STATE OF PLAY H.A.C. Fay and K.H. Halfpapp Queensland Horticulture Institute Queensland Department of Primary Industries PO Box 1054 Mareeba Queensland 4880 Abstract Fruitpiercing moths (Eudocima spp.) are important pests of tree fruit and vine crops throughout the summer rainfall areas of eastern and northern Australia. Effective control is currently limited to netting trees or crops, or bagging fruit. This is costly and not always desirable for the crop involved. Growers have been seeking an alternative control method, which is pest specific and rapidly deployable if fruitpiercing moths pose a threat. This paper reports on the results of a series of field trials designed to assess the type and effectiveness of baits containing synthetic feeding attractants for moth control in citrus. Sugared-agar baits containing highly volatile fruity esters (such as n-butyl acetate and methyl butyrate), aldehydes and an alcohol (together totalling 25 µl/25 g bait) were more attractive to the primary moth species than baits that contained esters only. When low volatility esters were substituted in the volatile mix, bait attractiveness was substantially reduced. The best baits attracted 85% of primary moths (mainly Eudocima fullonia [Clerck]) on baited citrus trees through to the first week of harvest. This then fell to 75%, if the presence of Eudocima materna (L.) is discounted (as it was the least responsive of the species attracted to the baits). Waxing the exterior of baits extended their field life by several days. Attempts to find other ways of presenting the synthetic feeding attractants, involving various types of dispensers, have proved unsuccessful. The implications of this work are discussed, as are the future research needs. Key words: citrus, fruit piercing moth, Eudocima spp, synthetic feeding attractants Introduction Fruitpiercing moths, Eudocima spp. (Lepidoptera : Noctuidae), feed on ripening fruit after piercing the skin or rind with a strong, barbed proboscis (Tryon 1898). They attack a large range of tree and vine fruits, the main ones being lychees, guavas, mangoes, kiwi fruit, citrus, papaws, persimmons and carambolas (Fay 1996). In Australia, fruitpiercing moths are common along the eastern seaboard, primarily in coastal areas, and across the northern half of the continent. The most important species are E. fullonia, E. materna, and Eudocima salaminia (Cramer). Control of fruitpiercing moths cannot be achieved with conventional insecticide coversprays as (1) insufficient contact of the moth with the fruit denies knockdown, and (2) fruit are attacked too close to harvest to achieve an adequate withholding period. Effective control has only been attainable through expensive exclusion techniques, such as crop netting or bag covers. Light protection systems suppress moth feeding by 60-70% at low-moderate population levels (Fay and Halfpapp 1994), but they require a substantial investment in infrastructure for a single pest, which varies in severity annually and locally. Evolving IPM systems in several affected crops, such as citrus and papaws, require that a targeted control method is developed for fruitpiercing moths which can be rapidly deployed if, and when, these pests are a problem. Development of baits incorporating feeding attractants commenced with electroantennography on E. fullonia and E. materna in which differential responses to individual synthetic fruit odour components signified the general fruity esters as prominent (Fay and Halfpapp 1994). Combinations of odour components, particularly where two different chemical groups were involved, produced electroantennogram responses greater than expected from the observed responses to the same components offered singly. Combinations of chemicals were developed, based largely on the major odour components of kiwi fruit. These were incorporated into sugared-agar baits for flight cage and then field studies of attractiveness. Experiments with E. fullonia, in which baits contained different ratios of specific fruit volatiles (esters, aldehydes and alcohols) corresponding to the odours produced at different stages of ripeness, indicated that 51 ripe ‘fruit’, as opposed to mature or very ripe, were preferred. The current paper reports on experiments conducted in citrus to evaluate bait type and effectiveness, and alternate volatile dispensing methods, as the basis of a lure-toxicant system for fruitpiercing moths. Materials and Methods Study site Trials were conducted in a citrus orchard of 29 trees on Southedge Research Station, 15 km northwest of Mareeba. The orchard consisted on Clementine and hybrid mandarins, and included the varieties Marisol, Denules, Page, Sunburst, Burn, Nova, Fina and Imperial. Fruit started to mature from early March, with most trees completing fruiting by May. Trees usually produced crops of several hundred fruit, but this depended on variety. Experimental baits and dispensers Agar baits were developed as artificial ‘fruits’ to test combinations of volatile attractants. These baits, when attacked by moths, showed distinct feeding damage for both primary and secondary species. Baits could withstand rain and be standardised across treatments. A full description of the preparation of both the agar baits and dispensers can be found in Fay and Halfpapp (1997). Attractant volatiles A full list of the volatile chemicals used in this study, their purity and suppliers, can be found in Fay and Halfpapp (1997). Experiments In all experiments, baits and/or dispensers were placed in citrus trees about three hours before dark. They were attached to the outer branches at 1.6-1.8 m above the ground. If fruitpiercing moth activity was monitored, it was undertaken 30 minutes after sundown for a period of 1.5 hours, with all primary piercers on baits/dispensers and fruit recorded, as well as the major secondary species. The morning after the exposure, agar baits were assessed for primary and secondary damage (i.e. feeding marks). 1994 - These experiments compared the field attractiveness of baits containing esters only against those that contained 80% esters plus aldehydes and an alcohol. They were conducted in the crop from early colour break to the fully ripe stage. Baits were either attached to 2-m poles adjacent to a fruit tree or to the tree itself. There were five treatment trees and associated poles per occasion. Counts of fruit were undertaken on treatment trees for estimates of moths/bait and moths/fruit. A subjective assessment of the proportion of the crop, which was ripe, based on colour, was also made. 1995 - Baits were placed in a citrus crop from 16 March to 26 April, on a single night each week. They were only attached to trees, but procedures were otherwise the same as in 1994. Three different treatment baits were used per tree, with each tree receiving two baits of each treatment. There were five baited trees. Three different volatile combinations were used: (1) four high volatility esters, three aldehydes and an alcohol, (2) two high volatility esters, an aldehyde and an alcohol, and (3) three low volatility esters, three aldehydes (as in 1) and an alcohol (as in 1). Esters contributed the same proportion of the volatile mix in each treatment. Each 25 mL bait contained 25 µl of volatile attractants. During the course of the trial, fruit on treatment trees were rated for colour from 0-100 (20 fruit/tree) and periodic measurements of Brix made to ascertain sweetness. 1996 - Dispensers were made up as described in Fay and Halfpapp (1997), and included 0.25% methomyl as the insecticide in the attractant/sugar solutions. Dispensers for each of the two treatments (1 and 2 in 1995) were placed in 10 trees, with two of each treatment/tree. Dispensers (and fruit) were observed for moth activity for two hours on the night they were placed out and again 4-5 days later. They were left on trees for one week before replacement. They were deployed between 14 March and 3 April. Through the experimental period, 20 randomly selected fruit from the treatment trees were assessed weekly for % Brix (using an Atago hand refractometer) and pH (to project to acidity) (after Kilburn and Davis 1959). 52 1997 - Standard agar baits (containing three esters [80%], three aldehydes and an alcohol) were compared against similar baits which had been dipped for a few seconds in hot paraffin wax. Immediately prior to immersion, the volatile mix, in proportion by volume to that in the baits, was added to the hot wax. There were 10 waxed and 10 non-waxed baits on each of 10 trees. Baits were assessed for moth damage the morning after the first night’s exposure. Night monitoring of moth activity took place on the third night only to ascertain bait durability. This experiment was duplicated on another occasion except that an additional treatment incorporated 2% glycerine and 2% paraffin oil into the agar baits before waxing. Results 100 100 90 90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 Average % of fruit showing colour Average % of baits attacked 1994 - Figure 1 shows the average % of the different bait types attacked by moths over a number of occasions in relation to crop maturity. For both bait types, the proportion of baits attacked by primary species declined as the crop ripened, whereas the number of attacks by secondary species increased. However, the baits that contained esters, aldehydes and an alcohol were substantially more attractive to the primary species than those that contained esters only. 0 17/03/1994 24/03/1994 Primary & Secondary spp. (all baits) Primary spp.(80% esters+ald.+alc.) 20/04/1994 Primary spp.(ester only baits) Fruit showing colour Figure 1. Percentages of different agar baits attacked by primary and secondary moth species in citrus in 1994 53 100 Period in which harvest would normally occur Baits damaged by primary moth species (%) 90 Bait type 1 (7 compts.) 80 Bait type 2 (4 compts.) Bait type 3 (low volat.esters) 70 60 50 40 30 20 10 0 16/03/1995 22/03/1995 29/03/1995 5/04/1995 12/04/1995 20/04/1995 26/04/1995 Figure 2. Percentages of three different bait types damaged by primary moth species in 1995 1995 - Figure 2 indicates the changes in the % of each bait type attacked by primary moths over the crop ripening period. More of the seven component baits were attacked through the first half of the experiment and overall than the other two bait types. Baits containing low volatility esters were far less attractive to moths (p < 0.05) than the other two bait types. E. fullonia was the most numerous species during the first half of the experiment, with E. materna dominating during the second half. E. salaminia and E. jordani were also recorded. Moth numbers peaked during the first half of April. 54 WITH E. MATERNA 100 Period in which harvest would normally occur 6 90 No. of moths observed 80 5 70 60 4 50 3 40 30 2 20 1 10 0 Remaining fruit showing colour(%) 7 0 WITHOUT E. MATERNA 7 Period in which harvest would normally occur 6 90 No. of moths observed 80 5 70 60 4 50 3 40 30 2 20 1 10 0 Remaining fruit showing colour (%) 100 0 16/03/1995 22/03/1995 29/03/1995 Moths on baits 5/04/1995 12/04/1995 20/04/1995 26/04/1995 Moths on fruit (baited trees) Remaining fruit showing colour Figure 3. Moths observed on baits and on fruit on baited trees when E. materna is a) included, and b) excluded, from the data Figure 3 compares the numbers of primary moths observed on all baits with those on fruit on baited trees for (a) all primary piercing species; and (b) all primary piercers excluding E. materna. When all species were considered in the comparison there was no difference in the numbers of moths on baits or fruit (F = 0.68, P = 0.427), whereas the exclusion of E. materna revealed significantly more moths (75%) on baits (F = 5.76, P = 0.037). During the first three weeks of the trial, 85% of moths (all species) occurred on baits and only 15% on fruit on baited trees. 55 1996 - No moths were observed on dispensers in the field, although moths were known to respond to them in the laboratory. When dispensers were replaced by agar baits containing insecticide, only a single moth was recorded on them. This suggested that the insecticide used in both may have reduced the attractiveness of the volatile mix. 1997 - There was no significant difference in the numbers of waxed and standard baits pierced on the first night they were exposed. However, on the third night, five moths (all E. fullonia) were observed on the waxed baits but none on the standard ones. During the three-night exposure, 90% of waxed baits were heavily pierced compared with 50% of the standard baits (with some piercing). Glycerine/paraffin supplemented baits received slightly more piercings and attracted more moths than the non-supplemented waxed baits. Discussion Baits that contained esters, aldehydes and an alcohol were more attractive to primary fruitpiercing moth species than those that contained esters only. Ester type also appears important to the level of attraction, with the highly volatile general fruity esters, such as n-butyl acetate and methyl butyrate, eliciting a greater response than the less volatile and more specific esters. Despite five different primary piercing moths being recorded on the baits, E. materna was by far the least responsive. The trials indicated that E. materna may prefer a different volatile mix, or that the ratios of the existing components in the best baits may need adjusting. This is despite electroantennogram studies showing negligible difference in the responses of E. materna and E. fullonia. Finding a bait attractive to E. materna will be essential for growers in the Northern Territory and Western Australia where this species predominates. Nevertheless, the work described here has shown that baits attracted 85% of moths on baited citrus trees to the first week of harvest. This then declined to 75%, if E. materna is excluded. Waxing the exterior of baits extended their field life by several days without compromising attractancy. Baiting appears to offer better moth control than the light repulsion technique, but falls short of the 100% control achievable by netting. Baits containing synthetic feeding attractants work on the principle of short- range attraction in competition with fruit odour, thus decoying fruitpiercing moths away from the crop. In this respect, they are not like fruit fly lures which can attract males over hundreds of metres. As a consequence, these fruitpiercing baits are unlikely to attract moths to a crop, which are additional to those that would arrive anyway. Prior to moth numbers peaking in a crop it would seem unnecessary to bait every tree to significantly reduce damage. Probably every third or fourth tree would be sufficient. At peak harvest time bait deployment might need to be increased to every tree, or every second tree, to be effective. However, previous studies (Fay and Halfpapp 1993) have shown that most fruitpiercing moth damage in large square blocks of trees occurs in the two outermost rows, which may mean confining the baits to the orchard periphery. In some isolated situations the use of attractant baits prior to crop harvest may cause a localised reduction in the fruitpiercing moth population, negating any potential problem. Synthetic feeding attractants could also be used in fruitpiercing moth traps without the need for specialised dispensers or insecticides. This method would provide some control and allow moth activity levels to be ascertained prior to crop susceptibility. Further research is required to; (1) confirm a suitable toxicant to incorporate in the baits, (2) examine additional means to moderate volatile loss and extend bait field life (3) adjust the bait components to make them more attractive to E. materna, and (4) test the technique’s effectiveness in crops other than citrus. The technique may then offer fruit growers a specific means to control fruitpiercing moths at minimal cost. Acknowledgment This research was supported by a grant from the Horticultural Research and Development Corporation. The Queensland Horticulture Institute has authorised publication of this paper. 56 References Fay, H. (1996). Fruitpiercing moths on citrus - a perspective including control developments. QDPI Farmnote, Agdex 220/622. Fay, H.A.C. and K.H. Halfpapp (1993). Non-odorous characteristics of lychee (Litchi chinensis) and carambola (Averrhoa carambola) pertaining to fruitpiercing moth susceptibility. Aust. J. Exp. Agr. 33 : 327-331. Fay, H.A.C. and K.H. Halfpapp (1994). Feeding attractants and light barriers as control measures for fruitpiercing moths - progress in their evaluation. Pp. 89-96. Fifth Workshop in Tropical Entomology, Townsville, 1-5 July 1991. Fay, H.A.C. and K.H. Halfpapp (1997). Attractants and repellents for fruitpiercing moths. HRDC Final Report, Project No. CT 301, 30pp. Kilburn, R.W. and T.T. Davis (1959). The taste of citrus juice. 2. Citrate salts and pH. Proc. Florida State Hort. Soc. 72 : 271-276. Tryon, H. (1898). Orange-piercing moths. Qld. Agr. J. 2: 308-315. 57 INTRODUCTION OF TETRASTICHUS BRONTISPAE FOR CONTROL OF BRONTISPA LONGISSIMA IN AUSTRALIA Keith Halfpapp Queensland Horticulture Institute Mareeba Queensland 4880 Abstract The palm leaf beetle, Brontispa longissima Gestro, was recorded from Moa Island in Torres Strait in 1911, and the Australian mainland at Cooktown in 1976 (Fenner 1987). Since then it has been recorded at Darwin 1979 (Fenner 1987), Broome 1989 (Strickland et. al. 1991) Cairns 1992, Innisfail 1995 and Kuranda 1997 (unpublished data). The parasite Tetrastichus brontispae (Ferriere) (Hymenoptera: eulophidae) was introduced into Australia from Western Samoa in 1982 (Fenner 1987) and host specificity testing was carried out by Bill Yarrow, QDPI Indooroopilly, in that year. Four thousand parasites were then released in the Darwin area during the first half of 1983, and 200 parasitised beetle pupae were sent to QDPI Mareeba for release at Cooktown. No field recoveries were made after these releases. Ted Fenner is of the opinion that these releases failed because the parasitoid colony came from a laboratory culture. In May 1984, a second introduction of the parasite was sourced this time from ORSTOM in New Caledonia. An estimated 220 F2 parasites were released in Darwin and later recovered on the 21 June 1984. Parasitism levels in pupal samples collected to November 1984 ranged from 060.3% with an average of 20.7%. Palm leaf beetle was found in Redlynch and Freshwater (northern suburbs of Cairns) in November 1992 (unpublished data). It is suspected that the insect had been present in the area for approximately 12-18 months before its detection. At this time T. brontispae could not be sourced from the Northern Territory. An application was made to re-import the parasite, sourced from Guam, in May-June 1994. The original material was quarantined at Sherwood and then mass reared at QDPI Mareeba. Tetrastichus brontispae was released at 11 sites in the Cairns area, four sites at Thursday Island and three at Seisa and Bamaga in late 1994/95. Parasitised pupae were also sent to the NT and WA. Field bred parasites were recovered from most release sites on the mainland within one month. In October 1995 the author was seconded to work on the PFF program and further work on palm leaf beetle control was suspended until the appointment of Sharyn Foulis in January 1997. Sharyn again established a culture of the parasite, obtaining material from the NT as well as local sites. Her followup surveys conducted in 1997 recovered T. brontispae from two sites in Cairns and two in Cooktown. During 1997-98, a further 14 releases were made in the Cairns/Cooktown/Innisfail area. Parasites have been recovered from four of these release sites in 1997-98. It has been established that parasites moved, unaided, 12 km from a release site north of Cooktown to a population of palm leaf beetle at Hopevale Mission and also 4 km distance in the Cairns area. Using the Climex Program and modified weather data for Mareeba, (Skarratt et al. 1995) a distribution map predicts that B. longissima would survive across the Top End, Cape York and the length of the Queensland coast and into the coastal area of northern New South Wales. This distribution is similar to that predicted when weather data for Noumea is used in the program. Only in one instance in the North Queensland release program, at Endeavour Falls, has T. brontispae demonstrated an ability to greatly reduce populations or limit the expansion of palm leaf beetle. Successful release is dependent on the initial release of large numbers of the parasite. 59 References Fenner, T.L. (1986). Biological control of Brontispa longissima in the Northern Territory. In Proceedings of the Fourth Workshop on Tropical Agricultural Entomology, p. 111-115. Strickland, G.R., Shivas, R. and Young, S.J. (1991). The use of the green muscadine fungus Metarhizium anisopliae, to control Palm Leaf Beetle, Brontispa longissima, (COLEOPTERA : CHRYSOMELIDAE/ in Broome. In Proceedings of the Workshop in Tropical Entomology No. 5. P. 134-137. Skarratt, D.D., Sutherst, R.W. and Marywald, G.F., (1995). Climex for Windows Version 1.0. CSIRO and CRC for Tropical Pest Management, Brisbane. 60 BIOLOGICAL CONTROL OF PALM LEAF BEETLE, BRONTISPA LONGISSIMA (GESTRO) (COLEOPTERA: CHRYSOMELIDAE) WITH THE WASP PARASITOID, TETRASTICHUS BRONTISPAE (FERRIERE) (HYMENOPTERA: EULOPHIDAE) IN DARWIN Deanna Chin and Haidee Brown Entomology Branch Department of Primary Industry and Fisheries GPO Box 990 Darwin NT 0801 Abstract The eulophid parasitoid, Tetrastichus brontispae (Ferriere) was imported from New Caledonia and released in Darwin to control the palm leaf beetle, Brontispa longissima (Gestro). Coconut palms at six sites were monitored before and after the introduction of the parasitoid. B. longissima first appeared in Darwin in 1979. T. brontispae was first introduced into Darwin to control B. longissima in 1982. The initial introduction did not establish, but a new introduction in 1984 established for five years and then died out (Fenner 1992). T. brontispae was reintroduced into Darwin in 1994 and was established in moderate numbers for two years. Between October 1994 and March 1997, the beetle damage to the coconut palms at the release sites was reduced by 20%. The parasitoid was established in higher numbers at sites that were irrigated with overhead sprinklers. After November 1996, the numbers of T. brontispae had diminished and it could not be collected from any of the release sites or nearby areas. The climate in the Top End of the Northern Territory may be responsible for the parasitoid’s failure to establish, as it is probably suited to a milder tropical climate. Key words: biological control, Brontispa longissima, Tetrastichus brontispae Introduction In Darwin, coconut palms (Cocos nucifera) are grown for ornamental purposes and are commonly planted in backyards, parks and amenity areas. B. longissima is a major pest of coconut palms in Darwin, Palmerston and rural areas of the Top End of the Northern Territory. Occasionally, severe damage can be seen on royal palms (Roystonea regia (H.B.K.) O.F. Cook), and moderate to minor damage occurs on some ornamental palms such as carpentaria palm (Carpentaria acuminata Becc.), fishtail palm (Caryota mitis Lour.) and clumping golden cane palm (Chrysalidocarpus lutescens H. Wendl.) (Jones 1984). A host list has been included in the DPIF Agnote by T. Fenner (1989). In Australia, B. longissima is also found in northern Queensland, north Western Australia (Broome) and, overseas, it ranges from Java eastward to New Caledonia, Tahiti, Western Samoa and American Samoa. B. longissima was first found in Darwin in 1979. An eradication campaign was attempted, but was abandoned in 1981. Chemical sprays are useful in controlling B. longissima, but, as the palms reach their mature height, (over 10 metres), spraying is often difficult. To reduce reliance on chemical sprays and to control B. longissima in tall coconut palms, the parasitic wasp T. brontispae was introduced. T. brontispae is endemic to Indonesia (Java, Nusa Tenggara) and Irian Jaya (Kalshoven 1981). T. brontispae attacks the late larval stages and 1-2 day old pupae of B. longissima. T. brontispae was first introduced into Darwin and Palmerston in 1982 from Western Samoa, but failed to establish. An introduction of New Caledonian stock in 1984 established successfully and survived for about five years (Fenner 1992). In 1994, the Entomology Branch of the Department of Primary Industry and Fisheries re-introduced the parasitoid as field releases in Darwin. T. brontispae was imported from New Caledonia to the Department of Primary Industries in Brisbane, Queensland where it was reared and screened through quarantine. After the wasps were reared for three generations, a small consignment of the parasitoid was sent to Berrimah Research Farm, where it was mass reared for release in Darwin. 61 Study Sites A total of six sites were selected in the Darwin City Council amenity areas (Table 1). These sites included picnic areas, street plantings, experimental plantings and parks. All of the sites were located near the foreshore except site 4. Table 1. Study sites selected for monitoring B. longissima infestation levels in coconut palms. Site No. Location Description 1* Cullen Bay Overhead irrigation in dry season 2* Mindil Beach Overhead irrigation in dry season 3 East Point Reserve Overhead irrigation in dry season 4* Coconut Grove No irrigation in dry season and was burnt during the 1996 dry season 5* Nightcliff Beach Overhead irrigation in dry season 6* Near Nightcliff Swimming Pool No irrigation in dry season * Release site for T. brontispae Materials and Methods Monitoring B. longissima damage levels Damage by B. longissima to the new leaves of coconut palms was assessed from March 1993 to December 1998. From March 1993 to March 1997, coconut palms at the sites were monitored once every 6-8 weeks (the interval between production of new fronds) and from March 1997 to December 1998 the sites were monitored once every six months. Monitoring commenced 12 months prior to the release of T. brontispae. At each inspection, 12 coconut palms were selected randomly and given a damage rating from 1 to 5 to assess the level of damage (Table 2). The damage rating was assessed by an observer (standing at ground level) viewing the newly opened frond and assigning a damage rating. Table 2. Ratings used to assess the level of B. longissima damage to coconut palms Damage rating Beetle damage to the first opened leaf (%) 1 1-20 2 20-40 3 40-60 4 60-80 5 80-100 Laboratory rearing T. brontispae was mass reared in a laboratory at 25-30ºC and 60-90% relative humidity, located at Berrimah Research Farm. The rearing containers were clear cylindrical plastic tubes (45 x 110 mm) with fine mesh lids and bases. The wasps were fed honey smeared onto cotton wicks 62 and placed on the outside of the fine meshed lid. Twice a week, the wasps were provided with late instar larvae or early pupae of B. longissima (field collected) for ovipositing. Each rearing tube contained between 100-300 adult wasps and was provided with 10-30 beetle pupae. The mean number of wasps reared per beetle pupa was 10, but ranged from 4-18. To maintain genetic diversity in the culture, T. brontispae was also established in coconut palms in the gardens at Berrimah Research Farm and on a regular basis, infested B. longissima pupae were added to the laboratory cultures. Field releases Field releases were carried out between October 1994 and April 1996. Tubes of parasitised B. longissima pupae were taken to field sites at weekly, fortnightly or monthly intervals. Each tube was placed in a coconut palm, at the base of an infested leaf spear, with the lid removed so that the wasps could disperse at emergence, generally within seven days. Overall, about 13,000 T. brontispae were released in the selected sites between October 1994 and April 1996 (Table 1). During field monitoring, samples of B. longissima pupae were collected and reared in the laboratory to see if T. brontispae was established at the sites. Results The effectiveness of T. brontispae in controlling B. longissima was determined by comparing the damage to the coconut palms before and after release. Results of monitoring collected between March 1993 and March 1997 are summarised in Table 3. Within two years after the release of T. brontispae, there was a 20% decrease in the level of B. longissima damage at three of the release sites (site 1, 2 and 5). Site 4 did not have a damage assessment before release, however, from general observations of the site just before release, the damage level was estimated to be about 60%. At site 6, there was no change in the beetle damage after release. In the control site (site 3), damage levels fluctuated from 0-60%. Between October 1994 and November 1996, parasitised beetle pupae were collected from all release sites, indicating that T. brontispae was established. After November 1996, T. brontispae could not be detected from any of the monitoring sites; however, it was still present in the coconut palms at Berrimah Research Farm. Table 3. The level of B. longissima damage on coconut palms at the study sites before and after the release of T. brontispae (based on results collected between March 1993 and March 1997) Site No. Level of B. longissima Infestation Before release of T. brontispae After release of T. brontispae Change in infestation level since release of parasitoid 1*+ 4 3 Decreased 20% 2*+ 3 2 Decreased 20% 3 1 2 Increased 20% 4* - 3 - 5*+ 3 2 Decreased 20% 6* 2 2 No change * Release site for T. brontispae + Irrigated site 63 Discussion T. brontispae initially decreased B. longissima damage by 20% at sites that were watered by overhead irrigation, however, it was not able to establish in the long term. The improvements in the growth of the palms and the decrease in B. longissima damage was noticeable in the wet season when the palms were vigorous in growth and the wet weather provided favourable conditions for the survival and increase in numbers of T. brontispae. The effect on B. longissima was further compounded by the presence of a fungal pathogen, Metarhizium anisopliae (Metschnikoff) (the green muscardine fungus) (Banks 1979). This pathogen is prevalent during the wet season (rarely seen in the dry season) and is effective in killing all stages of B. longissima (D. Chin unpublished data) T. brontispae and its effect on controlling B. longissima have been studied in the Pacific Islands. In some cases T. brontispae was shown to be effective (Chiu and Chen 1985 and Stapley 1980) and in others it only provided minor control (Hollingsworth et al. 1985 and Voegele 1989). T. brontispae was released in North Queensland in 1994 and has been established successfully and may provide control in the long term (K. Halfpapp pers. comm.). In Darwin the main limiting factor restricting the establishment of T. brontispae appeared to be due to climatic conditions. T. brontispae did not survive well during the prolonged dry season (which is about six months). This strain may be more suited for establishment in a climate similar to New Caledonia which has evenly distributed rainfall (Rudloff 1981). Strains of T. brontispae from the drier areas of Indonesia or Papua New Guinea may be more suited to the climate in the Top End. There were higher numbers of T. brontispae (reared from B. longissima pupae) collected from overhead irrigated sites compared to non-irrigated sites. In addition, adults of T. brontispae were observed on coconut fronds at Berrimah Research Farm in a garden that was watered daily with overhead sprinklers. A small amount of competition may have occurred between T. brontispae and natural pathogens and predators of B. longissima. Both T. brontispae and M. anisopliae were more prevalent during the wet season. Since M. anisopliae affects all stages of B. longissima, this could reduce the availability of hosts for T. brontispae. M. anisopliae may have hindered the wasp’s ability to produce sufficiently large populations, which may have assisted its survival over the dry seasons. The green tree ant, Oecophylla smaragdina (Fabricius) appeared to be an important predator of all stages of B. longissima. O. smaragdina was found in large numbers on coconuts and as a result may have also limited the availability of hosts for T. brontispae. Fenner (1996) noted that populations of B. longissima were reduced in coconut palms that were inhabited by tree frogs, geckoes and a species of ant, Tetramorium simillimum (F. Smith). Conclusion T. brontispae was able to establish, and exert limited control of B. longissima in Darwin for two years. Sites that were irrigated with overhead sprinklers provided more suitable conditions, however, the climate was still unsuitable for long term establishment. The parasitoid appeared to be adapted to climate that has lower temperatures and more consistent rainfall throughout the year. Acknowledgements We thank Heather Wallace for assisting with the laboratory rearing and field monitoring. We thank also Stuart Smith and Ted Fenner for helpful advice and suggestions on the release program. We are grateful to Graham Young, Glenn Bellis, Stuart Smith and Ted Fenner for reviewing the manuscript. 64 References Brady, B.L.K. (1979). Metarhizium anisopliae. Descriptions of Pathogenic Fungi and Bacteria. Commonwealth Mycological Institute Kew, Surrey, England. No. 609. Chiu, Shui-chen and Chen, Bing-huei. (1985). Importation and Establishment of Tetrastichus brontispae, a Parasitoid of the Coconut Beetle in Taiwan. TARI. Special Publication No.19, pp 12-13. Fenner, T.L. (1992). Biological Control of Brontispa longissima in the Northern Territory. Tropical Agricultural Entomology, 4th Workshop, Darwin NT, 10-15 May 1987. Pp 111-115. Fenner, T.L. (1996). Palm Leaf Beetle. Department of Primary Industry and Fisheries. Agnote no. 371. Hollingsworth, R., Meleisea, S. and Iosefa. (1988). Natural Enemies of Brontispa longissima (Gestro) in Western Samoa. Alafua Agriculture Bulletin. Pp 41-45. Jones, D. (1984). Palms in Australia. Reed Books Pty Ltd. Australia. Pp 110, 173 and 184 Kalshoven, L.G.E. (1981). Pests of Crops in Indonesia. (Van der Laan, P.A. and Rothschild, G.H.L. Trans.) PT Ichtiar Baru-Van Hoeve Indonesia (pp 449, 582-583). Stapley, J.H. (1980). Coconut Leaf Beetle (Brontispa longissima) in the Solomons. Alafua Agriculture Bulletin. Vol. 5. No. 4. Pp 17-22. Rudloff, W. (1981). World-Climates with tables of climatic data and practical suggestions. Wissenschaftliche Verlagsgesellschaft mbH Stuttgart, Germany. Pp 563 and 579. Voegele, J.M. (1989). Biological Control of Brontispa longissima in Western Samoa: An Ecological and Economic Evaluation. Agriculture, Ecosystems and Environment. Elsevier Science Publishers B.V., Amsterdam. 27: 315-329. Yarrow, W.H.T. (1982). A Report on the Host Specificity Testing for Tetrastichus brontispae (Ferriere), a Parasite of the Palm Leaf Beetle, Brontispa longissima (Gestro). Department of Primary Industries Qld. 8 pp. (Unpublished). 65 EGG PARASITOIDS OF FRUITSPOTTING BUGS (AMBLYPELTA SPP.): POTENTIAL AND LIMITATIONS OF MANIPULATIVE RELEASES H.A.C. Fay1, S.G. De Faveri1 and R.K. Huwer2 1 Queensland Horticulture Institute Queensland Department of Primary Industries PO Box 1054 Mareeba Queensland 4880 2 CSIRO Entomology GPO Box 1700 Canberra ACT 2601 Abstract Three different egg parasitoids of the banana-spotting bug, Amblypelta lutescens lutescens (Distant), were found in Mareeba in 1992. The potential to mass-rear and strategically release one or more of these parasitoids (Anastatus sp., Ooencyrtus caurus Huang and Noyes and Gryon sp.) to reduce bug breeding success was examined in preliminary laboratory, glasshouse and field experiments. Fruitspotting bugs produce eggs too slowly for mass-rearing parasitoids and so the utilisation of alternate host eggs was investigated. O. caurus could be reared on the eggs of some other coreids, e.g. Mictis spp.. The Anastatus sp. accepted a wide range of alternate hosts, with the eggs of a Neomantis sp. and an Opodiphthera sp. producing goodsized wasps in favourable sex ratios. In a series of glasshouse experiments which involved spotting bug eggs fixed to the leaves of papaw and macadamia plants, no parasitism was recorded after releases of either O. caurus or the Gryon sp. However, up to 84.4% parasitism of bug eggs by the Anastatus sp. was attained under similar circumstances. Parasitism levels were enhanced by increased numbers of bug eggs, releasing larger numbers of Anastatus or with adult bugs present near the eggs. Releases of Anastatus sp. and O. caurus into crops in the field in the absence of bug activity resulted in negligible parasitism of deployed bug eggs. However, confining pairs of adult bugs with eggs on carambola trees resulted in maximum parasitism levels of 22-50% following small-scale releases of Anastatus sp. Investigations should now look at larger scale releases of Anastatus and elucidate bug breeding habits, movement patterns and crop associations, to determine the potential value of these parasitoids for fruitspotting bug control. Introduction The banana-spotting bug, A. l. lutescens, and fruitspotting bug, Amblypelta nitida Stål, are severe pests of a large range of tropical and subtropical tree and vine crops through Queensland, parts of NSW, the Northern Territory and Western Australia (Ironside 1981, Donaldson 1983). Nymphs and adults pierce plant growing points, flowers or developing fruits and nuts. The damaged area of fruits becomes blackened and splits or cracks, and growth terminals dieback. Premature fruit and nut fall is common (Waite 1990). Effective control in susceptible crops is usually achieved through regular applications of endosulfan. Prior to 1992 natural enemies of spotting bugs were poorly known in Australia, with only some general predators (ants, spiders and assassin bugs) and the tachinid parasitoid, Pentatomophaga bicincta de Meijere, reported by Ironside (1981). In October and November 1992 three different wasp parasitoids were recorded attacking the eggs of A. l. lutescens in Mareeba (Fay and Huwer 1993). This parasitoid complex included an Anastatus sp. (Eupelmidae), O. caurus (Encyrtidae) and a Gryon sp. (Scelionidae). Overall percentage parasitism ranged from 37.5 – 91.6% over 3 sites, with the Anastatus sp. dominating. These parasitoids, and the levels of parasitism measured, were similar to those recorded for related coreids in other parts of the world (Phillips 1941, Greve and Ismay 1983, Oswald 1990). Exploration of native egg parasitoids through their mass-rearing and strategic release has been mooted to enhance control of dasynine bugs (Oswald 1990). This approach has been employed in China since the 1960s where Anastatus spp. are mass-released seasonally to control lychee stink bug (Tessaratoma papillosa Drury) in lychees and longans (Jianzhong 1990). Greater than 90% parasitism is achieved with this approach and costs are a third to a quarter of insecticidal 67 control. In Australia, there is increasing pressure on researchers to find alternative ways of combating fruitspotting bugs. This pressure is being driven by (1) the threat to withdraw or limit the use of endosulfan, (2) a lack of approved alternate chemicals in the large range of minor tropicals affected by spotting bugs, (3) large losses to Amblypelta in some organically-grown crops, ( 4) good progression of IPM and a reduction in insecticide use in a number of crops susceptible to spotting bugs, and (5) some progress with bug pheromone identification. Fruitspotting bugs are mobile pests, which may breed significantly on native host plants from where they invade nearby orchards. Routine chemical control practices ensure limited breeding in susceptible crops, as bugs in all developmental stages are killed by endosulfan. However, spotting bugs enter orchards continually while crops are developing, and eggs would be laid through the life of a crop. If insecticidal sprays are skipped, or the interval between sprays is too great, eggs and nymphs will survive and damage levels will increase. The integration of parasitoid releases with insecticidal sprays may reduce spray frequency while maintaining control levels. As parasitoids are mobile, commercial growers could release them in the vicinity of adjoining properties from where bugs may invade after breeding undisturbed, and over which the growers would normally have no control. This paper undertakes a tentative examination of the potential and limitations of rearing and releasing native egg parasitoids against A. l. lutescens. Materials and Methods Laboratory procedures Bug colony – Adults of A. l. lutescens were collected from a range of crop and ornamental hosts throughout the Mareeba district during the latter half of 1991 to establish the spotting bug colony. Complete details of how adults and nymphs were maintained can be found in Fay and De Faveri (1997). Parasitoid colonies – The wasp parasitoids described earlier originated from sites in Mareeba. Details of how individual species were held and reared are described by Fay and De Faveri (1997). Alternative host eggs – Spotting bugs have a low oviposition rate (2 to 4 eggs/day) which limits the degree to which their eggs can be used for parasitoid production. Releases of parasitoids in some commercial crops would require hundreds of thousands of wasps, and to produce these would necessitate either an alternative mass reared host or artificial eggs. A range of eggs of other insect species was tested as possible alternatives for rearing the parasitoid species discussed here. Eggs were obtained from insects collected in the field, and were selected on the basis of their size, known susceptibility to the parasitoid genera involved, and/or likelihood of laboratory colonisation. The eggs of six alternative host species were assessed for the Anastatus sp. and three for O. caurus. Gryon spp. are generally considered host specific and were not included in these tests. Initial attempts were undertaken to rear some of the alternative host species by caging/containing adults in the laboratory, providing appropriate food and/or offering host plant material for oviposition. Data were obtained (in most cases) on average egg weight, the sex ratio of wasps produced, male and female wasp size, and observations made on parasitism levels and parasitoid development times. Experiment details Parasitoid assessments were undertaken in the confines of a large glasshouse or in crops in the field. a) Glasshouse The dimensions of the building were 10 x 20 x 5.5 m. Plants (including tea, mangoes, papaws, macadamias, bananas and passionfruit) were held on 1 m high benches beneath an overhead watering system. Macadamias and papaws were grown in clusters about 5 m apart in the centre of the building. 68 Egg exposure method – Bug eggs were glued singly to 5 mm2 pieces of 0.4 mm thick rigid card using a small drop of wood glue. Eggs were always freshly collected for use in experiments, and were only exposed to parasitoids for 3-4 days before replacement or experiment termination. Experiments – A series of 10 experiments were undertaken; the complete details are described in Fay and De Faveri (1997). Different numbers of parasitoids (21-50/release) were used in the various experiments and released 0.5 or 10 m from the bug eggs. The length of the experimental period varied from 6-10 days. In two experiments, adult bugs were confined with the plants holding the bug eggs. In all situations there were three plants with eggs/treatment, and for all macadamia/papaw experiments plants either had some leaves coated with streaks of honey or did not (i.e. there were four treatments per experiment). Where experiments had more than one release of parasitoids they occurred three days apart. Where bugs were placed with plants, a single pair was confined in a gauze-covered capsule amongst the treatment group of plants containing the eggs. All parasitoids were more than one day old at release, with the Ooencyrtus and Gryon released at emergence sex ratios, and the Anastatus at a male:female ratio of 1:4. b) Field Sites – Experimental releases of parasitoids were conducted at two sites on the Atherton Tableland, at Walkamin Research Station and on the property of I. Steinhardt, between Mareeba and Walkamin. No insecticidal sprays were used on the experimental crops (papaws, macadamias and carambolas) at Walkamin Research Station and only fungicides were applied to the papaws at Steinhardt’s. Experimental plots at Walkamin Research Station included a row of 100 papaws separated into blocks of ten by single three year-old macadamia trees. Adjacent to this were eight year old carambola trees (originally 4 blocks of 20 trees) of five different varieties. The Steinhardt site contained about 800 widely spaced large papaws separated intermittently by small avocado trees. Experiments – A summary of the field assessments can be found in Fay and De Faveri (1997). There were three exposures of eggs during each experiment at Walkamin Research Station, and 11 at Steinhardt’s. The numbers of trees with bug eggs varied. Experiments were conducted over 9-11 days at Walkamin, where only Anastatus were released in batches of 25. At Steinhardt’s over the 60 day experimental period (21 October – 20 December) a total of 856 parasitoids (716 O. caurus and 140 Anastatus) were released. Adult bugs were confined near the bug eggs on one occasion at Walkamin. Parasitoids were released at 2-4 day intervals at Walkamin Research Station and at nine evenly spaced intervals at Steinhardt’s over 8½ weeks. All parasitoids were several days old and fed at release. At Walkamin, eggs were used singly on card, whereas at Steinhardt’s, five eggs/card were employed. Eggs placed on trees were retrieved by marking appropriate leaves with flagging tape. During the last two weeks of the Steinhardt experiment, eggs were placed on trees 15 m apart in a ring, with parasitoids released from the centre. Results Laboratory rearing Table 1 compares the average egg weight and sex ratio of Anastatus reared in spotting bug eggs against those for a range of alternative host eggs. The lunar moth, Ochrogaster lunifer Herrich-Schäffer, produces large egg masses coated in deciduous scales. While an average 66.4% of its eggs were parasitised by the Anastatus sp. in the laboratory, for every male only 0.46 females were produced. However, for the eggs of most other bugs, the gum moth and the mantid, a large female bias was recorded for emerging parasitoids, with no deterioration in wasp size. In particular, eggs of the Neomantis sp. (which lays rafts of 40 or so eggs) and those of the gum moth (which lays several hundred eggs at a time), were particularly well accepted by the Anastatus sp. and produced wasps of a very similar size to those produced from spotting bug eggs. O. caurus could be reared on the eggs of Mictis spp. (Coreidae), producing 12.07 wasps/egg, but it was far less receptive to alternate hosts than the Anastatus sp. 69 Table 1. A comparison of the suitability of a range of host eggs for rearing the Anastatus sp Average egg weight (g x 10-3) Parasitoid sex-ratio M:F Coreidae 1.3 1 : 2.22 Mictis profana Coreidae 4.5 1 : 5.50 Austromalaya sp. Pentatomidae 2.0-3.0 1 : 1.00 Neomantis sp. Mantidae 1.5 1 : 4.67 Ochrogaster lunifer Thaumetopeidae 1.9 1 : 0.46 Opodiphthera sp. Saturniidae 1.5-2.5 1 : 4.55 Host species Host family A. l. lutescens Glasshouse experiments In experiments where O. caurus and the Gryon sp. were released adjacent to bug eggs in papaws and macadamias, no parasitism was recorded on each occasion. 100 90 80 % parasitism 70 60 50 40 30 20 Two releases of wasps 10 Two releases (increased nos. bug eggs) 0 0 3 7 10 Days after initial wasp release Figure 1. Percent parasitism of fruitspotting bug eggs by Anastatus sp. released near-by in a glasshouse In experiments in which releases of the Anastatus sp. occurred adjacent to bug eggs, the proportion of parasitism attained seven days after the initial release reached an average maximum of 79% and 82% (or 84.4% if infertile eggs are excluded) (Figure 1). One experiment included increased numbers of bug eggs and parasitism levels were sustained at an average 80.7% over three egg collections. 70 Where Anastatus sp. were released 10 m from plants containing the bug eggs, parasitism was reduced an average 20% compared to the levels measured above. Figure 2 shows the parasitism levels recorded under different experimental regimes. A maximum level of 61.0% was recorded when adult bugs were contained near eggs on plants. 70 60 % parasitism 50 40 30 20 Single release of wasps 10 Two releases (with increased wasp nos.) Two releases (with FSB adults on plants) 0 0 3 6 9 Days after initial wasp release Figure 2. Percent parasitism of fruitspotting bug eggs by Anastatus sp. released 10 m away in a glasshouse Field experiments In the initial field experiment at Walkamin Research Station, where eggs were placed on papaw and macadamia trees, no parasitism by Anastatus was recorded. Predation by ants accounted for 13.3% of eggs on macadamias and 4.4% on papaws. A follow-up experiment in papaws and carambolas again produced no parasitism. In a third experiment, where adult bugs were confined near eggs on trees, parasitism reached 22.2% on a single occasion in the release tree, and 50% in an adjacent tree. Parasitism averaged 8.1% over the experiment (Figure 3). Predation by ants averaged 5% of eggs. 71 60 Maximum % parasitism 50 Release tree Adjacent tree 40 30 20 10 0 0 3 7 10 Days from initial w asp release Figure 3. Maximum percent parasitism of fruitspotting bug eggs by Anastatus sp. released into a carambola orchard with adult bugs present The series of Ooencyrtus and Anastatus releases in papaws at Steinhardt’s, where a total 346 bug eggs were exposed over 60 days, resulted in 1.2% parasitism and 9.8% egg predation. No spotting bugs were observed or feeding damage recorded during this experiment. Discussion The glasshouse and field trials implied that O. caurus and the Gryon sp. are unlikely to be of value in releases against spotting bugs, although more work on them is justified. Most evidence from S.E. Asia and China (Zhicheng et al. 1988) supports the view that Anastatus spp. have more potential than many other parasitoids in heteropteran bug suppression. In the glasshouse experiments reported here, the Anastatus sp. achieved a maximum 84.4% parasitism of bug eggs, a level similar to that recorded in the field in China when large numbers of Anastatus were released against lychee stink-bug (Zhicheng et al. 1988). Over the entire series of experiments undertaken in the glasshouse, the maximum parasitism achieved by Anastatus per experiment averaged 63.3% ± 18.1%, a level which would impact on bug numbers if it could be translated to the field. Further indications from the glasshouse experiments were that increased levels of or sustained parasitism by Anastatus occurred when (a) increased numbers of bug eggs were used, (b) increased numbers of parasitoids were released, and (c) spotting bug adults were present near the eggs on the plants. It is known that some parasitoids are attracted by pheromones produced by adult bugs (Aldrich et al 1984), which suggests that egg parasitoid host finding may be strongly linked to bug presence. Where relatively small numbers of parasitoids and bug eggs were used in the field in the absence of adult bugs, virtually no parasitism was recorded in our studies, irrespective of the crop or proximity of parasitoid releases to the host eggs. This implies that parasitoids could not be released as a prophylactic measure preempting bug activity in crops. Bugs would need to be in the crop already, or nearby, to justify releases. The fact that low-moderate levels of parasitism were achieved by Anastatus when bug adults were present in the field is promising. These trials used limited numbers of parasitoids, with no more than 50 released per tree. In China, releases of 600 Anastatus per lychee tree are reported to achieve maximum levels of parasitism (Zhicheng et al. 1988). However, 72 considerably more knowledge is required on bug behaviour and ecology before parasitoids can be best utilised as control tools, or integrated into existing control systems. Mass-rearing fruitspotting bug parasitoids needs to overcome the problem that bugs lay too few eggs to form the basis of a production system for large scale wasp releases. In China, Anastatus is reared in artificial eggs, which contain a component of insect (usually Oak silkworm) pupal haemolymph (Li-Ying et al. 1988). Such a technique may not be practical in Australia, unless an acceptable type and source of insect haemolymph can be found. Ideally, artificial eggs requiring no insect component will be developed. The results described above indicate that for Anastatus at least, a large range of insect eggs was acceptable for parasitoid rearing, many producing wasps in favourable female:male ratios and good-sized individuals. Despite this, use of alternate host eggs would require considerable development effort, particularly into host rearing methods. Without ease of rearing and economy of production alternate host eggs will not provide the solution to large-scale production of parasitoids. The work described in this paper is of a very preliminary nature and more could be done to further explore this methodology. This might include: a) Evaluating the rearing of parasitoids in artificial eggs, whether developed locally or imported frozen from China; this is essential to trialing parasitoid inundative release potential on a large scale. b) Large releases of parasitoids, timed with expected bug incursions into crops and the integration of the technique with existing controls. c) Elucidation of the association between parasitoid activity and bug pheromone emission. d) Procurement of far more information on bug breeding habits, movement patterns and crop associations, to determine the limits of an inundative release method. Acknowledgments The Horticultural Research and Development Corporation, the Macadamia Industry and the Papaw Subcommittee of the Queensland Fruit and Vegetable Growers are all thanked for contributing funds to this project. The Queensland Horticulture Institute has authorised publication of this paper. 73 References Aldrich, J.R., Kochansky, J.P. and Abrams, C.B. (1984). Attractant for a beneficial insect and its parasitoids: Pheromone of the Predatory Spined Soldier Bug, Podisus maculiventris (Hemiptera:Pentatomidae). Environ. Entomol. 13: 1031-1036. Donaldson, J.F. (1983). The Australian species of Amblypelta Stål (Hemiptera:Coreidae). J. Aust. Ent. Soc. 22: 47-52. Fay, H.A.C. and Huwer, R.K. (1993). Egg parasitoids collected from Amblypelta lutescens lutescens (Distant) (Hemiptera:Coreidae) in north Queensland. J. Aust. Ent. Soc. 32: 365-367. Fay, H.A.C. and De Faveri, S.G. (1997). Egg parasitoids of fruitspotting bugs (Amblypelta spp.): Potential for their mass-rearing and strategic release. Hort. Res and Devel. Corp. Final Report, Project No. HG 308: 25 pp. Greve, J.E. van S. and Ismay, J.W. (Eds.) (1983). Crop insect survey of Papua New Guinea from July 1st 1969 to December 31st 1978. Papua New Guinea Agr. J. 32: 1-120. Huang, D.W. and Noyes, J.S. (1994). A revision of the Indo-Pacific species of Ooencyrtus (Hymenoptera:Encyrtidae), parasitoids of the immature stages of economically important insect species (mainly Hemiptera and Lepidoptera). Bull. Nat. Hist. Mus. Lond (Ent.). 63: 1-136. Ironside, D.A. (1981). Insect pests of macadamia in Queensland. QDPI Misc. Publ. 81007: 28 pp. Jianzhong, B. (1990). Progress of the biological control research and application in China. Proc. First Asia-Pacific Conf. Entomol., Chiang Mai, Thailand, 8-13 Nov. 1989. Pp 493-503. Ent. and Zool. Assoc. Thailand, Bangkok: 855 pp. Li-Ying, L. Wen-Hui, L., Chao-Shian, C., Shi-Tzou, H., Jai-Chi, S., Han-Sun, C. and Shu-Yi, F. (1988). In vitro rearing of Trichogramma spp. and Anastatus sp. in artificial “eggs” and the methods of mass production. In Trichogramma and other egg parasites, 2nd Int. Symp., Guangzhou (China), Nov. 10-15, 1986. Pp 339-352. INRA, No. 43, Paris. Oswald, S. (1990). Possibilities for the use of Ooencyrtus albicrus ((Prinsloo) (Hym., Encyrtidae) in an integrated pest management approach against the coconut bug Pseudotheraptus wayi Brown (Hem., Coreidae) in Zanzibar. J. Appl. Ent. 110: 198-202. Phillips, J.S. (1941). A search for parasites of dasynine bugs in the Netherlands Indies. Trans. R. Ent. Soc. Lond. 91: 119-144. Waite, G.K. (1990). Amblypelta spp. (Hemiptera:Coreidae) and green fruit drop in lychees. Trop. Pest Mgmt. 36: 353-355. Zhicheng, L., Zhiyong, W., Yiren, S., Jianfeng, L. and Wuhong, Y. (1988). Studies on culturing Anastatus sp. a parasitoid of Litchi Stink bug, with artificial host eggs. In Trichogramma and other egg parasites, 2nd Int. Symp., Guangzhou (China), Nov. 10-15, 1986. Pp 353-360. INRA, No. 43, Paris. 74 INSECT FAUNA SURVEYS ON RAMBUTAN, DURIAN AND MANGOSTEEN IN NORTH QUEENSLAND David Astridge Queensland Horticulture Institute Centre for Wet Tropics Agriculture P.O. Box 20 South Johnston Queensland 4859 Abstract Insect fauna surveys have been carried out on rambutan (Nephelium lappaceum L.), durian (Durio zibethinus Murr.) and mangosteen (Garcinia mangostana L.) in the coastal wet tropics of north Queensland. Preliminary fauna lists are presented of pest and beneficial arthropods for each crop. A brief description is also given of major pest damage and seasonal abundance. Introduction The three most common exotic fruits commercially grown in north Queensland are rambutan, durian and mangosteen. These are evergreen trees, with dense foliage, which have their origins in Southeast Asia. Production is expanding to meet increased market opportunities both locally and overseas. The exotic fruit industries of north Queensland have clear objectives to develop sustainable pest management practices. Unfortunately, very little is known about the insect fauna associated with these crops. The information gained from these surveys will identify major pests and beneficial insects associated with each crop. The aims of this project were to identify major and minor pests and associated beneficial species for all three fruit crops and to determine seasonal and crop phenology impacts on pest abundance. Future research will target individual insects responsible for causing economic damage with the aim to develop an integrated approach to pest control. The purpose of this paper is to give an overview of arthropods encountered thus far in the surveys and provide a brief description of damage symptoms and the seasonal abundance of major pests. Materials and Methods Five commercial orchards, which included all three crops, were monitored along the coast of north Queensland. Monitoring sites were located south west of Mission Beach (146º: 00'E, 17º: 55'S), South Johnstone (146º: 00'E, 17º: 37'S), Babinda, (145º: 55'E, 17º: 20'S), Deeral (145º: 57'E, 17º: 12'S) and north east of Mossman (145º: 56'E, 16º: 17'S). Monitoring started on 20 July 1997 and was carried out monthly. Insects were collected using several methods that included sweep nets, beating trays, manual collection, fogging and light traps using a stratified random sample of 30 trees per crop at each location. Each tree was visually assessed for obvious plant damage to the leaves, stems, flowers, fruit and trunk. Ten branches with new leaf flush were randomly checked for insect activity and plant damage. All specimens were recorded and unknown insects causing plant damage were collected for identification. Insects were identified with reference collections at the Centre for Wet Tropics Agriculture at South Johnstone or sent to taxonomists for identification. Results Major and minor pests have been assessed by visual field observations throughout the year. Minor insects were considered sporadic and rarely influenced yield decline or fruit quality. The major and minor insect pests of rambutan, durian and mangosteen are presented in tables 1, 2 and 3, respectively. Beneficial arthropods are presented in table 4. 75 Table 1. Major and Minor Pests of Rambutan grown in north Queensland ORDER FAMILY GENUS SPECIES LOCATION* ** FEEDING SITE Major Pests Hemiptera Hymenoptera Lepidoptera Lepidoptera Lepidoptera Lepidoptera Pseudococcidae Formicidae Noctuidae Noctuidae Pyralidae Tortricidae Planococcus Oecophylla Othreis Eudocima Conogethes * citri smaragdina sp. salaminia punctiferalis * T, I, B, D, M T, I, B, D, M T, I, B, D T, I, B, D T, I, B, D, M T, I, B, D fruit/stems/ flowers farms mealy bug Fruit piercing Fruit piercing Fruit boring Fruit boring Minor Pests Acarina Coleoptera Coleoptera Coleoptera Coleoptera Coleoptera Hemiptera Hemiptera Hemiptera Hemiptera Hemiptera Hemiptera Hymenoptera Hymenoptera Hymenoptera Lepidoptera Lepidoptera Lepidoptera Lepidoptera Thysoneoptera Tenuipalpidae Bostrichidae Chrysomelidae Chrysomelidae Chrysomelidae Curculionidae Coccidae Coccidae Coreidae Flatidae Pentatomidae Tessaratomidae Formicidae Formicidae Formicidae Noctuidae Tortricidae Noctuidae Noctuidae Thripidae Brevipalpus Sinoxylon Monolepta Rhyparida Rhyparida Myllocerus Icerya Pulvinaria Amblypelta Colgaroides Plautia Lyramorpha Tetramorium Pheidole * * * Achaea Oxyodes Selenothrips sp sp australis discopunctulata spp sp sp psidii lutescens lutescens acuminata affinis parens bicarinatum megacephala * * * janata tricolor rubrocinctus T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D T, I, B, D, M T, I, B, D B, D T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D T, I, B T, I, B I,D fruit/stems/flowers branches/stem flush leaves flush leaves flush leaves branches/stem fruit/stems fruit/stems fruit/stems/flowers stems/shoots/ fruit stems/shoots sap sucking stems Farms mealybug Farms mealybug Farms mealybug leaves leaves leaves leaves shoots/flowers/fruit * = To be identified. ** Codes for locations; Tully = T; Innisfail = I; Babinda = B; Deeral = D; Mossman = M. Table 2. Major and Minor Pests of Durian grown in north Queensland ORDER FAMILY GENUS SPECIES LOCATION* ** FEEDING SITE Major Pests Hemiptera Hemiptera Hymenoptera Coleoptera Lepidoptera Lepidoptera Coreidae Pseudococcidae Formicidae Chrysomelidae Pyralidae Tortricidae Amblypelta Planococcus Oecophylla Rhyparida Conogethes * lutescens lutescens citri smaragdina discopunctulata punctiferalis * T, I, B, D T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B T, I, B fruit fruit/stems/flower farms mealy bug flush leaves fruit borer fruit borer Minor Pests Coleoptera Coleoptera Hemiptera Hemiptera Hemiptera Hymenoptera Hymenoptera Hymenoptera Lepidoptera Lepidoptera Lepidoptera Cerambycidae Chrysomelidae Flatidae Pentatomidae Pentatomidae Formicidae Formicidae Formicidae Noctuidae Noctuidae Tortricidae * Prosoplus Monolepta Colgaroides Ancanthidiellum Accarana Tetramorium Pheidole * Autoba * * Sp. australis acuminata souefi australica bicarinatum megacephala * versicolor * * T, D T, I, B, D, M T, I, B, D, M T, I, B, D T, I, B, D T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M Trunk girdler flush leaves stems/shoots/fruit stems/shoots/fruit stems/shoots/ fruit Farms mealybug Farms mealybug Farms mealybug leaves/flowers leaves leaves * = To be identified. ** Codes for locations; Tully = T; Innisfail = I; Babinda = B; Deeral = D; Mossman = M. 76 Table 3. Major and Minor Pests of Mangosteen grown in north Queensland ORDER FAMILY GENUS SPECIES LOCATION** FEEDING SITE Major Pests Hemiptera Hemiptera Hymenoptera Thysoneoptera Coreidae Pseudococcidae Formicidae Thripidae Amblypelta Planococus Oecophylla Selenothrips lutescens lutescens citri smaragdina rubrocinctus T, I, B, D T, I, B, D, M T, I, B, D, M T, I, B, D, M fruit fruit/stems/ flowers farms mealy bug fruit/shoots Minor Pests Acarina Coleoptera Coleoptera Coleoptera Coleoptera Hemiptera Hemiptera Hymenoptera Hymenoptera Hymenoptera Tenuipalpidae Chrysomelidae Chrysomelidae Chrysomelidae Chrysomelidae Coreidae Flatidae Formicidae Formicidae Formicidae Brevipalpus Geloptera Monolepta Rhyparida Rhyparida Amblypelta Colgaroides Tetramorium Pheidole * spp miracula australis caeruleipennis clypeata nitida acuminata bicarinatum megacephala * T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I T, I, B T, I, B, T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M stems/shoots leaves/shoots leaves/shoots leaves/shoots leaves/shoots fruit stems/shoots/fruit Farms mealybug Farms mealybug Farms mealybug ** Codes for locations; Tully = T; Innisfail = I; Babinda = B; Deeral = D; Mossman = M. Table 4. Combined Beneficial Insects and Spiders of Rambutan, Durian and Mangosteen grown in north Queensland ORDER FAMILY GENUS SPECIES LOCATION* * HOST Araneida Araneida Araneida Araneida Araneida Coleoptera Coleoptera Coleoptera Lycosidae Oxyopidae Heteropidae Salticidae Araneidae Coccinellidae Coccinellidae Coccinellidae Lycosa Oxyopes Holconia Opisthoncus Gasteracantha Amidellus Coelophora Coccinnella spp. spp. spp. spp. sp. ementitor inoequalis transversalis T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M Coleoptera Coleoptera Coleoptera Coleoptera Coleoptera Coleoptera Diptera Hemiptera Hemiptera Mantodea Neuroptera Odonata Odonata Odonata Odonata Odonata Odonata Odonata Odonata Coccinellidae Coccinellidae Coccinellidae Coccinellidae Coccinellidae Staphylinidae Dolichopodidae Pentatomidae Reduviidae several * Chrysopidae Coenagrionidae Coenagrionidae Coenagrionidae Isostictidae Isostictidae Protoneuridae Lestidae Megapodagrionidae Cryptolaemus Illeis Micraspis Micraspis Rodolia several * Psilopus Oechalia Pristhesancus several * Mallada Ischnura Agrionocnemis Agrionocnemis Austrosticta Isosticta Alloneura Austrolestes Austroargiolestes montrouzieri galbula lineola frenata sp several * sp. schellembergii plagipennis several * sp. fragilis argentea dobsoni fieldi simplex coelestina insularis aureus T, I, B, D, M T, I, B T, I, B T, I T, I, B, D T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D, M T, I, B, D T, I, T, I, B T, I, B, D, M T, I, B, D, M T, I, T, I, B, D, M T, I, B, D, M General predator General predator General predator General predator General predator ? ? moth eggs, thrips, scale mealy bug ? ? ? Icerya spp. (scale) General predator moth eggs General predator General predator General predator General predator General predator General predator General predator General predator General predator General predator General predator General predator * = To be identified. ** Codes for locations; Tully = T; Innisfail = I; Babinda = B; Deeral = D; Mossman = M. Discussion Citrus mealy bug (Planococus citri Risso) and the green tree ant (Oecophylla smaragdina Fabricius) are common pests to all three crops. Although green tree ants have been reported as being beneficial in some fruit crops in drier climates (Peng et al. 1995), this has not been the experience of fruit growers in the coastal wet tropics of north Queensland. Favourable conditions along with the abundance of food and shelter favour the build up of large populations of green tree ants. These insects are active farmers of the citrus mealy bug and encourage high populations and subsequent sooty mould growth on fruit and leaves. In rambutans, green tree 77 ants web or tie fruit together making it difficult to clean. The territorial behaviour of the green tree ant makes them aggressive if disturbed and large populations can pose difficulties for efficient harvesting by attacking fruit pickers. If green tree ant populations were managed by injecting nests with a suitable chemical they may act as biological control agents of other insect pests. Alternatively, selective insecticides could be used to control citrus mealy bug. The ladybird, (Cryptolaemus montrouzieri Mulsant) was an effective predator of the citrus mealy bug. Its juvenile stages mimic the citrus mealy bug and move freely within green tree ant populations without being attacked. This predator has the potential to play an important part in the biological control of the citrus mealy bug in north Queensland as does the parasitoid, Leptomastix dactylopii (Howard) (Ceballo et al. 1998). Other minor pests in common with all three crops include swarming leaf beetles (Rhyparida sp., Monolepta (Monolepta sp.) and plant hoppers (Colgaroides acuminata Walker). A number of other beneficials have been previously recorded from many of the major pests mentioned for rambutans, durian and mangosteen (Ceballo et al. 1998 and Smith et al. 1997). The efficacy of biological control with these beneficials need to be quantified before consideration can be given to their use in developing integrated pest management programs. Plant Damage Descriptions and Seasonal Abundance for Major Pests of Rambutan Yellow peach moth (Conogethes punctiferalis Guenee.), an unidentified fruit borer and fruit piercing moths (Eudocima salaminia Cramer and Othreis spp.), cause damage to the fruit. The first two pests feed externally on green and mature fruit from December to March. On green, developing fruit the larvae also feed on pedicels causing premature fruit drop. As the fruit matures the larvae bore into fruit making it unmarketable. Fruit piercing moth adults feed on the fruit juices of maturing fruit and by damaging the skin allow the entry of secondary rots. These pests are most active from February to April. Citrus mealy bugs and green tree ants are active throughout the year with peak numbers occurring between December and April. Plant Damage Descriptions and Seasonal Abundance for Major Pests of Durian The Banana spotting bug (Amblypelta lutescens lutescens Distant) is most active in durians between October and January causing large numbers of fruit to drop prematurely as a result of bugs feeding on pedicels. On fruit, feeding causes cracked sunken areas as the surrounding tissue dies and the fruit continues to grow. The yellow peach moth and an unidentified fruit borer larva cause damage to the fruit as it matures between February and April. The larvae feed on the fruit surface between the spines covering themselves with frass. Black swarming leaf beetles (Rhyparida sp.) are active throughout the year with peak periods occurring between October and April. Under high populations, black swarming leaf beetles can severely set back tree development. Green tree ants and citrus mealy bugs are present on durian between November and March. Plant Damage Descriptions and Seasonal Abundance for Major Pests of Mangosteen Red-banded thrips (Selenothrips rubrocinctus Giard.) develop high populations during hot dry conditions between September and January. Thrips are very active in mangosteens from flowering and will continue feeding on developing fruit up to harvest. Most of the damage occurs as thrips feed on the fruit surface causing a russetting blemish, which results in downgrading or rejection of the fruit from the market. Fruit-spotting bugs cause damage to young developing fruit. Feeding sites develop into sunken lesions, which eventually crack as the fruit grows. Citrus mealy bugs and the green tree ants are a minor problem in mangosteens, possibly due to the thicker and more ridged leaves making them unsuitable for nest construction. 78 Acknowledgments I thank John Donaldson and Ross Storey for their assistance in preparing specimens and identification of insects, Bruno Pinese for general advice in the preparation of this manuscript and the Queensland Horticulture Institute for permission to publish. References Ceballo, FA., Papacek, D., and Walter, GH. (1998). Survey of mealybugs and their parasitoids in Southeast Queensland citrus. Australian Journal of Entomology 37, 275-280. Peng, R.K., Christian, K., and Gibb, K. (1995). The effects of the green ant, Oecophylla smaragdina (Hymenoptera: Formicidae), on insect pests of cashew trees in Australia. Bulletin of Entomological Research. 85, 279-284. Smith, D., Beattie, GAC., and Broadley, R. (Eds). (1997). Citrus Pests and their Natural Enemies: Integrated Pest Management in Australia. Information Series QI97030. Queensland Department of Primary Industries, Brisbane. 79 POTENTIAL OF USING COLONIES OF THE GREEN ANT, OECOPHYLLA SMARAGDINA (F.), TO CONTROL CASHEW INSECT PESTS Renkang Peng, Keith Christian and Karen Gibb Faculty of Science Northern Territory University Darwin NT 0909 Abstract The green ant, Oecophylla smaragdina, is a dominant predator of the main insect pests in cashew plantations in the Northern Territory, but its potential with respect to cashew yield must be documented before cashew growers will accept the ants as a biological control agent. In order to determine the appropriate technique and to provide the required documentation of the effectiveness of the technique, experiments and surveys investigating levels of green ant colonisation in relation to cashew yield were conducted in 1995, 1996 and 1997 at Wildman River Plantation, Northern Territory. Yield was closely related to the abundance of green ants in trees, and the yield pattern followed the following series: trees fully occupied by the ants > trees partly occupied by the ants > trees with low presence of the ants or trees without the ants. Yield was also positively correlated with ant colonisation level; 66 - 90% of the total variation in yield was accounted for by the level of ant colonisation. After insecticides were not used in cashew orchards, green ant colonisation levels increased from 0% to 80% over the first two or three years and then the level of colonisation oscillated between 52% and 70%. This resulted in a yield of 37 – 1,177 kg/ha at the level of 0 - 79% colonisation, respectively. However, experiments demonstrated that trees with ant colonies which were isolated from other colonies were much less damaged by the main insect pests and produced significantly higher yield than those with ant colonies which were not isolated. This is because fierce boundary fights between green ant colonies were eliminated. The level of ant colonisation in the isolated treatment was 100%, and the ant populations were high and stable, which resulted in yields of 1,943 to 2,683 kg/ha. It is suggested that cashew growers cannot rely on natural dispersal of the ants to control cashew insect pests and the isolation of colonies is a way to achieve high yield. The use of green ants as a key element to control insect pests of other tropical crops is also discussed. Introduction The green ant, Oecophylla smaragdina (Fabricius), is an effective predator, and it can significantly reduce the numbers of over 30 important insect pest species of many tropical crops (Way and Khoo 1992). The green ant can significantly reduce the damage levels of the main cashew insect pests, such as the tea mosquito bug, Helopeltis pernicialis (Stonedahl, Malipetil and Houston), the mango tip-borer, Penicillaria jocosatrix (Guenee), the fruit spotting bug, Amblypelta lutescens (Distant), the leaf-roller, Anigraea ochrobasis (Hampson), and the green vegetable bug, Nezara viridula (Fabricius) (Peng et al. 1995; 1997a, b; 1998). However, its potential with respect to cashew yield, which is the major concern of cashew growers, has not been fully demonstrated. In this paper, we present a series of large data sets to demonstrate that yield is related to different levels of ant colonisation. We also address the question of whether natural dispersal can be relied upon to obtain high yields and we describe the best way to maintain green ants at high levels in cashew orchards. Materials and Methods The work was done at Wildman River Plantation, which is in the wet-dry tropical area of the Northern Territory (12o64’S 131o87’E). Field survey Six field surveys were completed in three cashew blocks; block A with three surveys in 1995, 1996 and 1997, block B with two surveys in 1996 and 1997, and block C with one survey in 1997. Blocks A and B were 2.5 and 2.3 ha respectively, and the trees in these two blocks were 81 9 years old. Block C was 21 ha, and trees in this block were 6 years old. Blocks A and C had not received insecticides for two years, and Block B had not received insecticides for four years. Surveys were carried out during September and October of each year in blocks A and B. In each survey, every tree in the blocks was inspected. For each tree, the dominant ant species in the canopy, the number of ant trails on the main branches and the assessment of yield were recorded. Yield of each tree was assessed by the equation: Y = (N × T)/168; where Y is yield presented as kg/tree, N is the number of nuts/m2 in the shaded area under the tree canopy, T is a total shaded area of the tree canopy and 168 is the mean number of nuts/kg. We counted the number of nuts in each quadrat. Each quadrate was 1/4 m2, and two quadrates were done for each tree. Block C was divided to nine plots. These plots were surveyed in June 1997, and every tree in the plots was inspected. Yield was assessed in October 1997 for each plot. Recording procedures were the same as above. Field experiment The experiment examining the effect of the isolation of green ant colonies on insect pests was done in 1996 and 1997. The experiment had two treatments; isolation and no isolation. In 1996, in the isolation treatment, four green ant colonies, which occupied a total of 14 trees, were continuously isolated from April by chopping tree branches or twigs which linked to other trees occupied by other colonies. The boundary of each colony was determined by following the ant trails. In the no isolation treatment, the boundaries of 12 ant colonies were determined, and these occupied a total of 68 trees. Branches and twigs connecting ant colonies in this area were never cut. Trees used for this experiment were 10 years old and had not received chemical pesticides for 8 years. In 1997, in the isolation treatment, five colonies, four of which were the same as those used in 1996, occupied a total of 16 trees. These colonies were continuously isolated from March 1997. In the no isolation treatment, the boundaries of 9 ant colonies, which occupied a total of 33 trees, were determined but were not separated. The populations of the main insect pests in the two treatments were monitored from June to November. This is the period of flowering and fruiting for cashews, and it is also the period when the yield is the most susceptible to insect damage. Pest populations were quantified by recording the percentage damage of flushing shoots at the middle and bottom levels of the tree canopy. The damage by the main pest species was uniquely identifiable, and only fresh damage was recorded fortnightly at each sampling period. Ant fights and the abundance of the ant were monitored in the two treatment areas at fortnightly intervals from April to October by recording the number of ant fights on the colony boundaries and by counting the number of ant trails on the main branches of each tree. If more than 10 individual ants were counted walking along a main branch, the trail was assigned a value of 1. If less than 10 but more than one ant was counted walking along the branch, the trail was assigned a value of 0.5. At each monitoring occasion, 14 trees in each treatment were inspected. Data analysis One way ANOVA and the group t test were used to compare yield among different ant colonisation types and between treatments using the SYSTAT statistical package (Wilkinson 1990). Linear and polynomial regression analyses were used to explore the relationship between the level of ant colonisation and yield. Results Field survey The use of insecticides was stopped in block A in June 1995, and a survey in October 1995 showed that the block had no green ants. The yield from this block was only 0.2 kg/tree. Records of insect damage between June and September 1995 showed that 76% of the flower and flushing shoots were attacked by the tea mosquito bug and the fruit-spotting bug. In 1996, green ants started to colonise this block. From the distribution of green ants, meat ants (Iridomyrmex sanquineus) and black ants (Paratrechina sp) (Figure 1), green ants were more widespread than the other ants, and 37% trees had at least one green ant trail (Table 1). Trees with more than one green ant trail produced higher yield than trees with only one ant trail or trees with low numbers of green ants (Table 1). Trees without green ants had almost no yield. In 82 1997 after two years without insecticides, green ants had dispersed over most of the block (Figure 2), and 79% of trees had at least one green ant trail (Table 2). This resulted in a greater yield than in 1996 (Tables 1 and 2) and the same relationship between yield and ant colonisation type. In block B in 1996 after three years without insecticides, green ants had dispersed over most of the block (Figure 3). The proportion of trees with at least one green ant trail was 80% (Table 3). The relationship between yield and ant colonisation was the same as described above for block A. In 1997, the distribution of green ants was in patches (Figure 4), and the ants occupied only 63% of trees. However, the pattern of yield with respect of ant colonisation was the same as in 1996 (Table 4). In block C after two years without insecticides, the level of colonisation varied among plots from 32% to 79%, and the yield varied from 0.9 kg/tree to 6.4 kg/tree (Table 5). An analysis of the relationship between the level of ant colonisation and yield showed that plots with higher levels of ant colonisation generally produced higher yield, and that 66% of the total variation in yield was accounted for by the level of ant colonisation, Y = -3.01 + 0.13X; R2 = 0.66, (Fig. 5). Field experiment + P . . . . . . B . . + + . . . + . . . . P + . . . . . P + . . . . . . . . . . . . . . . . G P G . P . . P G P . . . . . . . . . . . G G . . . P P . . . P . G . . P G . . . . G . . . + P . P . . . . . P G . . . . . . . P G . . . . + G G G . + . . . . . P + G P P G P + . . . . . + . . . + + P + + P . . + G . G + + + . . . + + G . G . P P . P . . . . P G G G P P + + . . P . + . G G P G . + . . . . . P G G G P P . . P . . + . G . . . . . P + . . . + P P G G + G + + + P + + . . + + . . . . . . . . . . . G P G + + + + . . G + . . . . . . P P G . + P + G . + + . . . . + + . . + . . . . . . G P + . . . G . G P G G G . P + . . . . P . P G G G G + . + . + + P P G G G G P P + P . G P G . G G + + G + . G + + G G + G . G G G + . P . . P P P P G + + P P + G + G . G + P . + + G + G . . G P . G G G P P G + G P P P G + G G + G G . P . + G P + G G G . G P G P M + G G P G P . . + G G G = Trees fully colonised by green ants (> 1 ant trail/tree). P = Trees partly colonised by green ants (= 1 ant trail/tree). + = Trees with green ants, but no obvious ant trails. M = Trees with the meat ant, Iridomyyrmex sanguineus. B = Trees with the black ant, Paratrechina sp. . = Trees without ants. Figure1. Distribution of ant species in cashew block A in 1996 after one year without insecticides 83 The experiments showed that the average damage done during flowering and fruiting period by the tea mosquito bug, the mango tip-borer and the fruit spotting bug was 22.6%, 8.4% and 13.8%, respectively in the no isolation treatment. However, in the isolation treatment the damage caused by each of these insect pests was less than 1%. When comparing the abundance of green ants between the two treatments, it is found that the ants occupied 100% of trees in the isolation treatment, and each tree had 3 - 4 trails. However, in the no isolation treatment, the number of ant trails gradually decreased from 1.3 to 0.2 trails per tree and only 52 - 66% of trees were occupied by the ants. Table 1. The effect of the level of ant colonisation on the yield of cashews in block A in 1996 (one year without insecticides) Ant Oecophylla smaragdina O. smaragdina O. smaragdina Iridomyrmex sanguineus Paratrechina sp None 1 2 Colonisation type1 Number of trees Colonisation (%) G 91 21 2.80 + 2.84 a P + M 70 79 1 16 18 0.2 0.92 + 1.30 b 0.36 + 0.72 c 1.26 B . 1 197 0.2 45 0.00 0.14 + 0.48 c See Figure 1 for the explanation of the colonisation type. Means followed by the same letter are not significantly different at the 5% level. 84 Mean yield + SD (kg/tree)2 P G G G P G P + P G P . . . P P + G G P G P + G G G P P P . P G G . P G . . P B + P P P P G P P G G P P P P P P P P G P G P G G . . + G . P P G P P G + G G + P G G P + G P P P . P G + P P G + + P G G P G G P . G P . P P G + P + + P G P G . P P P G P P G G P G P G . P + P G G . . + . P G G P P G + P G G G B G P G G + P B . G P G P P + P + P + G P G . . G P G G P G G . G G P G + G P G G G G G P G P G G G G G P . . B G . . G P G P P P G G G G . P G . P . G G . G P P P P G + P P P P G P G G P P P G P . P G G G G G G P G + G G P P G G G P P . . G P P P . G G G + . G P P P P G P G G P P . G + . + G G G G G P G G G G P G G P P G G G P P P G G G P P G G P G + G G . . . . P P G G P P P + P P + G P G P P P G G + G G P . G M M P . . G G + G G G P . P G G P G G G G G G G + M G P + G G P G P P P P . P G G G G G G G G G G G + G G G P G + . P G + G M G G G G G G G G G P M G G M G G P G G M G G G G G G G G G + G G M G M G M G Figure 2. Distribution of ant species in cashew block A in 1997 after two years without insecticides. See Figure 1 for an explanation of the symbols in the figure Table 2. The effect of the level of ant colonisation on the yield of cashews in block A in 1997 (two years without insecticides) Ant O. smaragdina O. smaragdina O. smaragdina I. sanguineus Paratrechina sp None 1 2 Colonisation type1 G P + M B . Number of trees 211 151 38 10 4 47 Colonisation (%) 46 33 8 2 1 10 Yield + STD (Kg/tree)2 7.88 + 6.22 a 5.20 + 4.25 b 1.60 + 1.96 c 8.89 + 6.33 ab 3.24 + 4.11 d 2.25 + 3.44 cd See Figure 1 for the explanation of the colonisation type. Means followed by the same letter are not significantly different at the 5% level. 85 . . P P G G G G P P P P . . P P P G . . . P + G G P P P P G P . P G G G P P G P G G P + P + P P P P P . G G P P + G G P P G G G G G G P G + P G + G G P . G G G G G P P G P G P G G P G G G + P G G G G G . G . G G P P P G . G P G G . G P P + G G P G G G . P G G G G P P G P G + P G G G G + G P P P . G G G + P + G G P G P + P P + P G G M G G G + P . . + G + + + G G + G P P P P G . M G G + . + . . . G G + G G . G G G P G G G G G G G P G G P + G P G G + G G G G G G G P G P G G G G G P P . . P G G G G P P G + P P G G G G G . . P G G P P P . P G G + + G P + + G G G G + G P P . G G . + P G G G G + + G + G G G P P G G P G G G G G P . . P . G G + G + P P P G G P G G G G G P P G . + P P G G G G + G P . G + P P P G G P G G P G + G G P G G G P G G G G G G G G G + G P G G + + P P P G G P G P . G G . P P P G G G G G G G G M G G P G P . . M G G G P + P P G G G G G Figure 3. Distribution of ant species in cashew block B in 1996 after three years without insecticides. See Figure 1 for an explanation of the symbols in the figure The numbers of ant fights observed in the no isolation treatment were 13 in 1996 and 10 in 1997. Almost no ant fights were found in the isolation treatment. Trees in the isolation treatment produced a mean yield of 10.5 kg/tree in 1996 and 14.5 kg/tree in 1997, but in the no isolation the yield was only 4.6 kg/tree in 1996 and 3.9 kg/tree in 1997 (Table 6). This difference in yield was highly significant between the two treatments (Table 6). Table 3. The effect of the level of ant colonisation on the yield of cashews in block B in 1996 (three years without insecticides) Ant Oecophylla smaragdina O. smaragdina O. smaragdina Iridomyrmex sanguineus None 1 Colonisation type1 Number of trees Colonisation (%) G 221 52 7.66 + 3.92 a P + M 118 48 4 28 11 1 4.42 + 2.72 b 1.82 + 1.58 c 8.20 + 3.28 a . 38 9 1.62 + 1.44 c See Figure 1 for explanation of colonisation type. 2 Means followed by the same letter are not significantly different at the 5% level. 86 Mean yield + SD (kg/tree)2 P P P G G . P . . . G . P P P P P G P G G G P . + + B P G G + P P + . P P P P P + . P P P G G G . G + B P G G . P P P P G G G P G G + G P P G G + P . + . P G P + P P G P G P P P P P + P P G G P P . P . P G G + P P P M P G P P P P P P P M . P P P P + P + G G + P P P P G G G G G G P + . G P . . . . P G G P + P G P G G G + . + + P G + M P P . . + . + P B + . G P G P + G P + . M P P G + . + . + G G P P P G G P P G P G P P . P G P G + G P P G . + P . P G G . + G P P + + G . + G P P P . + G . + + P P P G G P + P + G . G G + P . . + P G P G G + . + + . . G . G P P + . P G G G P + . . . P P G P G G P + . P P P + P + P G P G + . . . . B G P P G P P . + + P + + P + G G P G P . . . . + B G P P P G + . . . . B P G + P P G + . G P P P G G . P G . G P G . G P G G P . M M P . B B . . G G P + . + G . P G G P G G G G M M G . M G G G G M + . . M M M G M Figure 4. Distribution of ant species in cashew block B in 1997 after four years without insecticides. See Figure 1 for an explanation of the symbols in the figure Table 4. The effect of the level of ant colonisation on the yield of cashews in block B in 1997 (four years without insecticides) Ant O. smaragdina O. smaragdina O. smaragdina I. sanguineus Paratrechina sp None 1 2 Colonisation type1 G P + M B . Number of trees 117 149 63 14 8 73 Colonisation (%) 28 35 15 3 2 17 Yield + SD (kg/tree)2 7.79 + 4.46 a 5.47 + 4.13 b 3.36 + 2.38 c 7.08 + 6.72 ab 1.03 + 0.99 d 3.39 + 2.34 c See Figure 1 for the explanation of the colonisation type. Means followed by the same letter are not significantly different at the 5% level. 87 Table 5. The relationship between yield and the level of green ant colonisation in block C Plot number Colonisation (%) Yield ± SD (kg/tree)1 Yield (kg/ha) 1 79 1,177 6.36 ± 3.81 a 2 63 577 3.12 ± 1.24 c 3 60 1,090 5.89 ± 3.05 b 4 57 971 5.25 ± 2.04 b 5 54 324 1.75 ± 1.89 d 6 50 635 3.43 ± 1.82 c 7 48 599 3.24 ± 1.18 c 8 33 167 0.90 ± 1.22 e 9 32 209 1.13 ± 0.93 e 1 Means followed by the same letter are not significantly different at the 5% level. Fig. 5 Yield (kg/tree) 8 6 4 2 0 0 20 40 60 80 100 The level of ant colonisation Figure 5. The relationship between the level of ant colonisation and yield of cashews Table 6. The effect of the isolation of O. smaragdina colonies on the yield of cashews Year Treatment No. of colonies No. of trees Mean diameter of tree canopy m Mean yield 1996 Isolation 4 14 7.3 + 0.6 a 10.5 + 2.1 a No isolation 16 68 7.0 + 1.1 a 4.6 + 3.0 b 1997 Isolation 5 16 6.9 + 1.5 a 14.5 + 4.9 a No isolation 9 33 7.2 + 2.8 a 3.9 + 2.9 b 1 Means followed by the same letter are not significantly different at the 5% level. Discussion Green ants are very susceptible to insecticides (Peng, unpublished data). When cashew trees were no longer sprayed with insecticides, green ants started to colonise the cashew orchard. The colonisation rate increased when ant populations were low, but the colonisation rate decreased when ant populations were high. In surveys of block A, two years after conventional sprays were stopped, the level of ant colonisation rose from 0% in 1995 to 80% in 1997 (Figures 1 and 2; Tables 1 and 2). When green ant populations were high in block B after three years without insecticides (Fig. 3; Table 3), the colonisation rate decreased from 80% in 1996 to 63% in 1997 (Figures 3 and 4; Tables 3 and 4). This is because when ant populations were high, fights between colonies took place, which resulted in a decrease in the ant populations (Peng et al. 1997a; 1998). Therefore, some trees, which were previously occupied by green ants became decolonised (Figure 3 vs Figure 4). The observed dynamics between green ant colonies confirms the results of our previous field surveys of an area of 2.4 ha cashews. When green ant 88 colonisation was low (57%) in 1993 (after six years without insecticides), the level of colonisation increased to 70% in the following year (Peng et al. 1997a). Thus, the level of ant colonisation rate was found to oscillate below 80% under natural conditions (Table 7). However, with isolation, ant colonisation was 100% in both years (Table 7), and ant populations in each tree were stable. This is because the fights between colonies were eliminated, and trees in each colony were fully occupied by the ants over the two years. By comparison, the level of ant colonisation in the no isolation treatment was only 66% in 1996 and 52% in 1997 (Table 7). Therefore, it appears that the elimination of fights between colonies is essential for high levels of ant colonisation, and the isolation of colonies is an effective way to solve this problem. Yield is closely related to both type and level of ant colonisation. A pattern of yield with respect to the ant colonisation type was established, that is trees fully occupied by the ants had a greater yield than trees partly occupied by the ants, and partly occupied trees had a greater yield than trees with low presence of the ants or trees without the ants (Tables 1 - 4). This demonstrated that trees with more abundant green ants produced higher yield. A summary of the information with respect to the level of ant colonisation in different blocks with trees of a similar age is shown in Tables 5 and 7, and the yield per hectare (one hectare contains 185 trees). Using the data from Tables 1 - 6 a regression analysis for block C (six-year-old trees) suggested that higher ant colonisation rate resulted in higher yield (Figure 5), and 66% of the variation in yield was explained by the level of ant colonisation. Table 7. Summary of the relationship between ant colonisation and yield with respect to different types of field management (raw data are from Tables 1, 2, 3, 4 and 6) Cashew block A A A B B Wildman* Wildman* Experiment Non-isolated Non-isolated Isolated Isolated * Tree age Years without insecticides Colonisation rate (%) Yield Type of 8 9 10 9 10 8 9 <1 1 2 3 4 6 7 0 37 79 80 63 57 70 37 156 1,087 1,042 997 ----- Natural dispersal Natural dispersal Natural dispersal Natural dispersal Natural dispersal Natural dispersal Natural dispersal 10 11 10 11 8 9 8 9 66 52 100 100 851 722 1,943 2,683 Natural dispersal Natural dispersal Colony isolation Colony isolation Data are from Peng et al. 1997a. In block A, after the regular use of insecticides was stopped, the level of ant colonisation increased from 0% to 79% and yield increased from 37 kg/ha to 1,087 kg/ha (Table 7). In block B, an 80% level of colonisation resulted in a yield of 1,042 kg/ha. In the following year, the level of colonisation decreased to 63% and yield was decreased to 997 kg/ha (Table 7). In the isolation experiment, the level of ant colonisation was only 66 - 52% in the no isolation treatment, and the resulting yield was 851 to 722 kg/ha. However, in the isolation treatment, ant populations in each tree within each colony were stable throughout the year and trees within each colony were all occupied by the ants (100% colonisation). This treatment gave a yield of 1,943 to 2,683 kg/ha (Table 7). A regression analysis (Figure 6) showed that 90% of the yield variation is accounted for by the level of ant colonisation (Y = 66.28 - 5.82X +0.27X2; R2 = 0.90). 89 Fig. 6 Yield (kg/ha) 3000 2000 1000 0 0 40 80 120 Level of ant colonisation (% ) Figure 6. The relationship between the level of ant colonisation and yield of cashews Six-year-old trees that were treated with 9 - 10 insecticide sprays between April and September, produced 1,230 kg/ha (Peng et al. 1997b). This yield is similar to the yield of 1,177kg/ha obtained in block C with six-year-old trees (Table 5) that had a level of 79% ant colonisation. Although no isolation experiments with six-year-old trees were done, the data obtained from 10 year old trees (Table 7) suggests that if the level of ant colonisation could be increased to 100% through the colony isolation process, the trees would have produced higher yields than those which were fully protected by insecticides. Therefore, it is suggested that the isolation of green ant colonies in cashew plantations can be very effective in increasing yield. Many tropical fruit crops share similar groups of insect pests with cashews, such as Helopeltis in cocoa, coffee and tea, Amblypelta and Noctuidae caterpillars in paw paw, citrus, rambutan, mango, jackfruit and macadamia etc. Therefore, green ants have the potential to be useful biological control agents in a variety of tropical crops. The distribution of green ants is very wide in southeast Asian countries from Sri-lanka, India, southern China through to the Solomon Islands and northern Australia (Cole and Jones 1948). They construct nests out of the leaves of over 100 species of native vegetation and many fruit trees in the tropical area (Majer 1990; Andersen 1992; Peng et al. 1997a). To make the best use of this ant on other tropical crops, three factors must be considered: (1) green ant population stability in orchards; (2) association with some homopteran insect pests; (3) aggressiveness to people. Green ants were traditionally used in southern China for many years, but were abandoned around 1960. The main reason is that farmers could not stabilise ant populations in their orchards and they had to repeatedly introduce the ants at least once a year. These activities were much more expensive than the use of insecticides at that time. During a visit to China, we found that farmers used to use the ants at the nest level, rather than considering the ant colonies. Our experiments on existing colony isolation in this study and transplantation of partial ant colonies with queens (Peng, unpublished data) suggest that using the ants at the colony level appears to be a way to stabilise ant populations. In this study, trees in the isolated colony area were never occupied by other ant species over the two year period because fights between colonies were eliminated and each colony was strong enough to guard its territory. However, in the non-isolated colony area, there were fights between colonies and as a result a few trees were not colonised. Later these trees were occupied by other ant species such as Crematogaster sp and Paratrechina sp. (Figures 3 and 4). Brown (1959) and Greenslade (1971) reported that competition between the green ant and other ant species is considered to be a problem in some areas of Papua New Guinea and the Solomon Islands. A study by Way and Khoo (1992) suggests that the use of an ant bait (Amdro) is effective in stabilising Oecophylla populations by killing or reducing other competitive ant species. Green ants have a close association with some homopterans such as scales, mealybugs, flatids and some species of aphids, and thus they do not control all the insect pests of tropical fruit 90 crops (Das 1959; Way 1963). However, the key question is whether the populations of scales and mealybugs exceed the economic injury level of the crops. There are a wide range of predators and parasitoids of scales, mealybugs, flatids and aphids. Chen (1962) suggested that green ants have a close association with the citrus scales, the mealybugs and the flatids in citrus groves in southern China, but they do not significantly affect the activity and population size of the main natural enemies of these insect pests. A positive association between green ants and coccids in citrus is also found in southern Vietnam, but coccids have never been identified as major pests according to the interviews with many citrus growers (Barzman et al. 1996). Preliminary observations suggest that when green ants are present, they do not affect homopteran predators and parasitoids resulting in very little damage to the cashew crop by homopterans (Peng, unpublished data). For a few crops like custard apple, mealybugs are the main insect pests during the fruiting period. An insect hormone pesticide known as Applaud is reported to be effective in controlling insect pests in the order Homoptera (Zhang 1989; Chen and Xiao 1990; Dai 1994). This pesticide is safe to many kinds of natural enemies (Zhang 1989; Chen and Xiao 1990). Green ants should not be affected by Applaud, but experiments are needed to confirm this. Aggressive behaviour is well known in green ants. Most importantly, ants disturb people during harvest (Way and Khoo 1989). Our experience over the last four years shows that green ants are aggressive to people, but their bites are not harmful, and usually the pain caused by ant bites lasts a few seconds and has no side effects after that. Ant aggression is not identified as a problem by citrus growers in family-managed holdings in China and Vietnam (Barzman et al. 1996; Xu Hongji, pers. comm.). In contrast, labour in big plantations is usually supplied by wage-earning workers who do not directly share the benefits of ant use and are therefore less likely to tolerate ant aggression. In this case, a way of decreasing the aggressiveness of the ants has to be found. Previous observations showed that the ants either went back to their nests or were less active when it was raining (Peng, unpublished data). Preliminary results using a plot of loofah vines showed that the number of ants on loofahs and the peduncles can be reduced by over 85% for 30 minutes after spraying with water. This gap of low ant activity can be even longer when the spray was carried out between 11:00 hrs and 15:00 hrs which is the period of least activity for green ants during a day (Peng, unpublished data). Therefore, fruit growers can use this method to reduce the disturbance from the ants during the period of harvest. Against this background, the green ant has much potential for use as a biological control agent in a wide range of tropical crops. The utilisation of the ant will be dependent on type of crops because each crop has its own complex of insect pests. Therefore, further research is needed to develop an integrated control program using the green ant as a key element for controlling insect pests in other tropical fruit crops. Acknowledgements The study was supported by the Rural Industries Research and Development Corporation and Northern Territory University. We thank L.N. Zhang for helping with field work and for helpful comments on the draft. We also thank Northern the Territory Land Corporation and Cashews Northern Territory for providing us with study sites. 91 References Andersen, A.N. (1992). The rainforest ant fauna of the northern Kimberley region of Western Australia (Hymenoptera: Formicidae). J. Aust. Ent. Soc. 31: 187-192. Barzman, M.S., Mills, N.J. and Nguyen, T.T.C. (1996). Traditional knowledge and rational for weaver ant husbandry in the Mekong delta of Vietnam. Agri. Hum. Valu. 13: 2-9. Brown, E.S. (1959). Immature nutfall of coconuts in the Solomon Islands. II. Changes in ant populations and their relation to vegetation. Bull. Entomol. Res. 50: 523-558. Chen, S. (1962). The oldest practice of biological control: the culture and efficacy of Oecophylla smaragdina Fabr. in orange orchards. Acta Entomol. Sin. 11: 401-407. Chen, X.F. and Xiao, Q. (1990). The efficient of using Applaud in tea plantations. China Tea 2: 25-27 (in Chinese). Cole, A.C. and Jones, J.W. (1948) A study of the weaver ant, Oecophylla smaragdina (Fab.). Am. Midl. Nat. 39: 641-51. Dai, X. (1994). The effect of Applaud on tea scales. Guei Zhou Tea, 4: 30-31 (in Chinese). Das, G.M. (1959). Observations on the association of ants with coccids of tea. Bull. Entomol. Res. 50: 437-448. Greenslade, P.J.M. (1971). Interspecific competition and frequency changes among ants in Solomon Islands coconut plantations. J. Appl. Ecol. 8: 323-352. Majer, J.D. (1990). The abundance and diversity of arboreal ants in northern Australia. BTROA 22: 191-199. Peng, R.K., Christian, K. and Gibb, K. (1995). The effect of the green ant, Oecophylla smaragdina (Hymenoptera: Formicidae), on insect pests of Cashew trees in Australia. Bull. Entomol. Res. 85: 279-84. Peng, R.K., Christian, K. and Gibb, K. (1997a). Distribution of the green ant, Oecophylla smaragdina (F.) (Hymenoptera: Formicidae), in relation to native vegetation and the insect pests in cashew plantations in Australia. Int. J. Pest Man. 43: 203-211. Peng, R.K., Christian, K. and Gibb, K. (1997b). Control threshold analysis for the tea mosquito bug, Helopeltis pernicialis (Hemiptera: Miridae) and preliminary results of the control efficiency by the green ant, Oecophylla smaragdina (Hymenoptera: Formicidae) in northern Australia. Int. J. Pest Man. 43: 233-237. Peng, R.K., Christian, K. and Gibb, K. (1998). Impact of native vegetation on cashew insect pests with particular reference to the most important pest - Helopeltis pernicialis. Rural Industries Research and Development Corporation, Canberra, pp. 70. Way, M.J., (1963). Mutualism between ants and honeydew-producing Homoptera. Ann. Rev. Ent. 8: 307-344. Way, M.J. and Khoo, K.C. (1989). Relationship between Helopeltis theobromae damage and ants with special reference to Malaysian cocoa smallholdings. J. Plant Prot. Trop. 6: 1-11. Way, M.J. and Khoo, K.C. (1992). Role of ants in pest management. Ann. Rev. Ent. 37: 479503. Wilkinson, L. (1990). SYSTAT: The System for Statistics (Evanston, IL: SYSTAT Inc.), pp. 677. 92 Zhang, J. W. (1989) A promising hormone insecticide (Applaud) to control the small leafhopper. Hunan Agriculture 5, 1-14 (in Chinese). 93 THE IPM OF SNAKE BEAN, VIGNA UNGUICULATA SSP. SESQUIPEDALIS, IN THE TOP END OF THE NORTHERN TERRITORY G. R.Young and L. Zhang Entomology Branch Department of Primary Industry and Fisheries GPO Box 990 Darwin NT 0801 Abstract Snake bean, Vigna unguiculata spp. sesquipedalis, is grown during the dry season in the Top End of the Northern Territory. The crop is grown, on trellises, under trickle irrigation. The most serious pests of snake bean are bean fly, Ophiomyia phaseoli, and two spotted mite (TSM), Tetranychus urticae. Bean fly is easily controlled with dimethoate, however TSM has rapidly developed resistance to the commonly used miticides. Snake bean crops protected with conventional miticides have produced beans for two to six weeks before increasing mite populations caused the abandonment of the crops. A snake bean crop was grown at Humpty Do, during 1995, using soft chemicals to control minor insect pests and the predatory mite Phytoseiulus persimilis to control TSM. P. persimilis was released at a rate of 100 per metre row, when TSM was first detected on the crop. TSM was successfully controlled by P. persimilis and as a result the productive life of the crop was extended to 12 weeks. One application, of either petroleum spray oil or fenbutatin oxide, prior to the release of P. persimilis enabled the predator to more quickly control TSM populations than without the spray. The use of P. persimilis to control TSM on snake beans would allow growers to plant two to three crops during the dry season as against the four to six crops currently grown using conventional miticides. Key words: Snake beans, Tetranychus urticae, Phytoseiulus persimilis, increased productivity. Introduction Vigna unguiculata ssp. sesquipedalis, yard long or snake bean, originated in tropical Africa and is now widely cultivated throughout the tropics as a green vegetable (Lazarides and Hince 1993; Verdcourt 1979). In the Top End of the Northern Territory snake beans, along with other tropical vegetables, are grown within 80 kilometres south and south east of Darwin. The annual rainfall varies from 1,600 mm in Darwin to 1,375 mm at Humpty Doo with 87 percent falling during the period November to March. For eight months of the year, evaporation exceeds precipitation. Mean maximum and minimum temperatures for Darwin during the wet season are 32.1 and 24.9°C respectively, while mean temperatures during the dry season, May to September, are 31.3 and 21°C (McDonald and McAlpine 1991). Snake beans are grown, on trellises, under trickle irrigation during the dry season. Production of snake beans has increased from 94 tonnes in 1994 to 136 tonnes in 1996 (PrimeStats 1995-96 and 1996). Most of the snake bean production is exported to the capital cities of the southern states. During the dry season the most serious pests are bean fly, Ophiomyia phaseoli (Tryon) (Diptera: Agromyzidae) and two spotted mite (TSM) Tetranychus urticae Koch (Acarina: Tetranychidae). Bean fly is controlled by two sprays of dimethoate within 10 days of germination (unpublished data), but TSM has remained a largely intractable problem due to the rapidity with which the mite became resistant to the recommended miticides; dicofol (Kelthane), fenbutatin oxide (Torque) and propargite (Omite). Crops protected with miticides produced beans for two to six weeks before increasing mite populations forced the abandonment of the crop (unpublished data). The Chilean predatory mite Phytoseiulus persimilis Athias-Henriot (Acarina: Phytoseiidae) was first reported from Australia by Goodwin and Schicha (1979) and has been used to control TSM 95 on strawberries in southern Australia (Goodwin 1990). As a result it was decided to try releases of the Chilean predatory mite to control TSM and consequently increase the productive life of snake bean crops. Materials and Methods Five rows each 40 metres in length were selected from a snake bean crop grown on a Humpty Doo property. The crop germinated on 29 April 1995 and was sprayed twice with 0.04% a.c. dimethoate within seven days of germination to control bean fly. After the initial applications of dimethoate, the crop was only treated with fungicides and insecticides which were known to be non-harmful to P. persimilis (Table 1) (Goodwin and Bowden 1984, Gough, undated). Table 1. Fungicides and insecticides used on the snake bean crop Chemical Treatment oxycarboxin Pirimicarb Bt as DiPel carbaryl Petroleum spray oil (PSO) as D-CTron Active concentration na 0.024% 0.01% 0.01% 2% Timing andNo. of applications 2 weeks post germination, 1 5 weeks post germination, 1 From flowering, 4 Spot sprayinfes ted plants Spot spray infested plants Reason Control rust Control Aphis craccivora Koch (Hemiptera: Aphididae) Lepidopteran larvae feeding on flowers and young pods Leaf feeding insects High TSM populations on individual plants. Eight leaves were sampled at random from, the lower, middle and top parts of plants in each row, making a total of 24 leaves sampled per row and 120 from five rows. Leaf samples were taken from six weeks post germination (pg) until the end of the crop. The leaves were returned to the laboratory and counts were made of the numbers of TSM per leaf and later, after the predator was released, the number of P. persimilis per leaf. Additionally, other predators found on the leaves were tested in the laboratory to ascertain if they would feed on TSM. Twenty thousand P. persimilis, purchased from a supplier in southern Queensland, were released on the five rows of crop within 10 days of the first detection of TSM. P. persimilis was released on the lower leaves. Observations were made on subsequent snake bean crops grown on the Humpty Doo property, during the 1996 and 1997 dry seasons, under the same conditions as described above, except that 2% DC-Tron® or fenbutatin oxide was sprayed once in the week before the introduction of P. persimilis. In all of these crops P.persimilis was released within 10 days of the first detection of TSM. Results Flowering started at six weeks and the first harvest of beans at eight weeks pg. Beans continued to be harvested the next 12 weeks until week 20 pg when the crop became senescent and was removed on 18 September 1995. There were few leaf-eating insects on the crop and as a consequence little carbaryl was used. TSM were first detected at seven weeks pg. The infestation appeared first on the lower leaves and then progressed to the upper leaves (Figures 1, 2 and 3). TSM populations peaked on lower leaves at week 12 pg, with 55% of leaves with greater than 5 TSM (Figure 1); on middle 96 leaves at weeks 13 and 14 with 17% of leaves having >5 TSM (Figure. 2); on top leaves at week 14 with 22% of leaves having >5 TSM (Figure 3). P. persimilis was released nine days after the initial detection of TSM. Populations of P. persimilis peaked at 15 weeks on lower, middle and upper leaves with means of 10.2, 15.4 and 14.6 per leaf, respectively. The percentage of leaves infested with P. persimilis at 15 weeks was 80, 67 and 65 for lower, middle and upper leaves respectively (Figures 1, 2 and 3). 97 80 12 10 70 8 60 50 6 40 4 30 20 2 10 No. of P. persimilis per leaf % lower leaves with TSM & P. persimilis % leaves with >5 TSM % leaves with P. persimilis No. of P. persimilis per leaf 90 0 0 8 9 10 12 13 14 15 16 17 18 Fig.1 PERCENT LOWER LEAVES WITH TSM & P. PERSIMILIS. NUMBERS OF P. PERSIMILIS PER LEAF % middle leaves with T.S.M >5 18 % middle leves with P. persimilis No. of P. persimilis per middle leaf 70 16 14 60 12 50 10 40 8 30 6 20 4 10 No. of P. persimilis per leaf % middle leaves with TSM & P. persimilis 80 2 0 0 8 9 10 12 13 14 15 16 17 18 70 % top leaves with T.S.M >5 60 No. of P. persimilis per top leaf 16 % Top leves with P. persimilis 14 12 50 10 40 8 30 6 20 4 10 2 0 0 8 9 10 12 13 14 15 16 Weeks Post Germination 17 18 FIG. 3 PERCENT TOP LEAVES WITH TSM & P. PERSIMILIS. NUMBERS OF P. PERSIMILIS PER LEAF 98 No. of P. persimilis per leaf % top leaves with TSM & P. persimilis Fig. 2 PERCENT MIDDLES LEAVES WITH TSM & P.PERSIMILIS. NUMBERS OF P. PERSIMILIS PER LEAF The predatory thrips Scolothrips ? sexmaculatus (Pergande) and larvae of Mallada basalis (Walker) (Neuroptera: Chrysopidae) were recorded feeding on TSM from weeks 10 to 16 pg. The larvae of the Cecidomyiid, Feltiella acarivora (Zehntner), were recorded feeding on TSM from week 15 pg onwards. F. acarivora was a voracious predator of TSM in the laboratory, with the last larval instar of the predator consuming up to 20 nymph and adult TSM per day. Iridomyrmex reburrus Shattuck (Hymenoptera: Formicidae) was observed foraging on the crop and removing pod boring larvae of an unidentified Lepidoptera. Observations on subsequent crops during the 1996 and 1997 dry seasons showed that P. persimilis could not survive on crops grown into the wet season. Populations of the predator would not persist in the humid conditions from October to March. Single sprays of either 2% DC-Tron® or fenbutatin oxide, applied four days before P. persimilis was released on beans, did not influence populations of the predator. In fact sprays of either of these chemicals retarded the rate of increase of TSM populations and allowed P. persimilis to control its prey in 20 days. F. acarivora appeared on crops some weeks after the TSM population had started to decline. Additionally, the following predatory mites were observed feeding on TSM; Amblyseius deleoni Muma and Denmark and Amblyseius longispinosus (Evans) (Acarina: Phytoseiidae) and an identified species (Acarina: Cunaxidae). S. ? sexmaculatus, M. basalis and the native predatory mites did not appear to exert any control over TSM populations in the absence of either P. persimilis or the use of miticides. Discussion Integrated pest management of TSM during the Top End dry season using P. persimilis allowed the harvest period of the snake bean crop to be extended from the usual four weeks to 12 weeks. P. persimilis controlled TSM in 35 and 49 days on lower and upper leaves, respectively. The progression of the TSM infestation from the lower to the upper leaves would indicate that sampling the lower leaves is sufficient when monitoring TSM populations on snake beans, prior to the introduction of P. persimilis. An initial introduction of one hundred P. persimilis per metre row appeared to be adequate to control TSM. Sprays with PSO and fenbutatin oxide applied prior to the introduction of P. persimilis, allowed the predator to more quickly control TSM than would have been the case without the sprays. Application of pesticides which are partially effective against TSM, but harmless to the P. persimilis, require further investigation. The naturally occurring predators, three species of predatory mite, S.? sexmaculatus and M. basalis did not appear to exert any influence on TSM populations. F. acarivora was an important predator but did not reach sufficient numbers quickly enough to control TSM. Gagné (1995) reported similar observations with other species of Feltiella. The appearance of the predatory ant I. reburuss shows the advantage of only using relatively soft pesticides. I. reburrus probably contributed to the mortality of leaf, flower and pod feeding insects. The use of P. persimilis in the IPM of TSM on snake beans would allow growers to plant two to three crops during the dry season as against four to six crops currently grown using conventional miticides. This should provide considerable economic benefits to growers. Acknowledgements Thanks are due to; Mr Kit Lim, Arnhem highway, Humpty Doo for allowing us to experiment with his crops; Dr Ray Gagné, Systematic Entomological Laboratory, USDA-ARS, Washington DC and Danuta Knihinicki, Agricultural Scientific Collections Unit, NSW Agriculture, Orange, NSW for identification of TSM predators; Deanna Chin, Entomology Branch, NT DPIF, for her comments on the manuscript. 99 References Gagné, R. J. (1995). Revision of the Tetranychid (Acarina) mite predators of the genus Feltiella (Diptera: Cecidomyiidae). Annals of the Entomological Society of America 88: 16-30. Gough, N. undated. Mites: Their biology and control on ornamental plants. Queensland Department of Primary Industries, QNIA/ QDPI. Goodwin, S. (1990). Seasonal abundance and control of spider mites (Tetranychidae) infesting commercial strawberries in coastal New South Wales. Journal of the Australian Entomological Society. 29: 161-166. Goodwin, S. and Bowden, P. (1984). Laboratory testing of the side effects of pesticides against predatory mites. In, Bailey, P. and Swincer, D. Eds. Proceedings of the Fourth Australian Applied Entomological Research Conference, Adelaide 24-28, 1984. South Australian Government Printer, Adelaide. Pp. 230-237. Goodwin, S. and Schicha, E. (1979). Discovery of the predatory mite Phytoseiulus persimilis Athias-Henriot (Acarina: Phytoseiidae) in Australia. Journal of the Australian Entomological Society. 18: 304. Lazarides, M. and Hince, B. (1993). Eds. CSIRO handbook of economic plants in Australia. CSIRO Melbourne Australia. 330 pp. McDonald, N.S. and McAlpine, J. (1991). Floods and drought: the northern climate. In Haynes, C. D., Ridpath, M.G. and Williams, M.A.J. Eds, Monsoonal Australia. Landscape, ecology and man in the northern lowlands. A. A. Balkema, Rotterdam. Pp. 19-29. PrimeStat, rural and fishing industries in the NT, 1995-96. Economics branch, NT Department of Primary Industry and Fisheries. Darwin. PrimeStats, rural and fishing industries in the NT, (1996). Economics branch, NT Department of Primary Industry and Fisheries. Darwin. Verdcourt, B. (1979). A Manual of New Guinea Legumes. Botany Bulletin No.11, Office of Forests, Division of Botany, Lae, Papua New Guinea. 645 pp. 100 IPM OF MELON THRIPS, THRIPS PALMI KARNY (THYSANOPTERA: THRIPIDAE), ON EGGPLANT IN THE TOP END OF THE NORTHERN TERRITORY G. R. Young and L. Zhang Entomology Branch Department of Primary Industry and Fisheries GPO Box 990 Darwin NT 0801 Abstract Melon thrips, Thrips palmi, is a polyphagous pest, which causes serious damage to capsicum, eggplant and Lebanese cucumber in the Northern Territory. Populations of T. palmi were highest during the last half of the dry season, July to September. Low populations during the wet season were thought to be the result of high prepupal and pupal mortality in saturated soil. T. palmi is highly tolerant of most of the commonly used chemical insecticides. Trials on eggplant showed that T. palmi can be controlled with potassium soap, a soft chemical, which does not affect the natural enemies of the pest. The most important natural enemies were the predators; Haplothips victoriensis, Deraeocoris sp. and Malada basalis. The use of persistent chemicals such as endosulfan and permethrin led to high thrips populations probably as a result of the insecticides killing thrips predators. It was concluded that T. palmi is an insecticide induced pest on eggplant in the Northern Territory. Key words: Thrips palmi, potassium soap, predators, persistent chemicals Introduction Thrips palmi Karny (Thysanoptera: Thripidae), melon thrips, was first described by Karny in 1925 from tobacco collected in Java (Bournier 1983). Prior to 20 years ago, T. palmi had only been recorded from Java and Sumatra (Bournier 1983). Since that time T. palmi has become widely distributed throughout South East Asia, China Japan, Taiwan, Hawaii, Papua New Guinea, a number of Islands of the South Pacific, Reunion, India, Pakistan, Africa, Central and South America, Caribbean islands and Florida USA (Johnson 1986; Waterhouse and Norris 1987; Walker 1994; Tsai et al. 1995). T. palmi was first detected from Australia in the Northern Territory (NT) in 1989 (Houston et al. 1991; Layland et al. 1994). Permanent populations of T. palmi are now established in the Darwin rural area, NT and, southeast Queensland (J. Layland, pers. comm.). The melon thrips is a polyphagus insect with a wide host range (Johnson 1986; Waterhouse and Norris 1987; Kajita, H., et al. 1996). In the NT a wide variety of vegetable crops are attacked including, beans (Vigna unguiculata and V. radiata), cucumber (Cucumis sativa), watermelon (Citrullus lanatus), bitter melon (Momordica charantia), squash and zucchini (Curcurbita pepo), pumpkin (Curcurbita maxima), capsicum and chili (Capsicum annuum) and eggplant (Solanum melongena) (Layland 1991). The most seriously affected crops are capsicum, eggplant and Lebanese cucumber (unpublished data). T. palmi feeds primarily on developing shoots and young leaves although flowers and young fruit are also attacked (unpublished data). Like most leaf feeding thrips, T. palmi punctures the epidermis, thrusts its mouthparts deep into the leaf tissue and sucks out the cell contents. The leaf surface develops a crinkly silvery appearance as a result of damage to the cells below the surface. Lightly infested plants show silver feeding scars on the underside of leaves, while more heavily infested plants show silvering and browning of leaves, stunting of terminal growth and scarred and deformed fruit. If eggplant, capsicum and Lebanese cucumber are attacked as seedlings, the plants remain stunted and eventually die (unpublished data). 101 T. palmi can reproduce sexually or by haplodiploid parthenogenesis (Kawai 1990). The female inserts her eggs into actively growing leaf tissue, flower buds or fruit. After hatching, two mobile and wingless larval stages feed on plant tissue. At the end of the second larval instar the larvae drop to the ground and burrow into the soil where they moult and become prepupae. This is followed by a further moult to the pupal stage. Winged adults emerge from the pupal stage, burrow to the soil surface and move to the plant hosts (Kawai 1990; Layland 1991, Layland and Brown, pers. comm.). High levels of soil moisture appear to predispose the prepupal and pupal stages to attack by fungi and bacteria (unpublished data). Tsai et al. (1995) reported the duration of life stages of T. palmi on eggplant at constant temperatures and 12:12 photoperiod (Table 1). Table 1. Mean development time for life stages of T. palmi on eggplant. (Adapted from Tsai et al. 1995) Temp. °C Egg Larva (I and II) 15 15.9 26 32 Prepup a+ Pupa Female Male 13.8 13.6 24.1 13.7 7.6 5.3 3.5 23.6 4.2 3.8 2.9 13.9 Mean Eggs Laid per Day Generation Time (days) Population Doubling Time (days) 18 58.4 17.8 12.6 29 29.1 7.4 11.1 20 16.6 4.5 Natural enemies of T. palmi recorded from the Northern Territory are the entomogenous fungus Verticillium lecanii (Zimm.) Viégas, which attacks newly emerged adults; the parasitoid Geotheana shakespearei Girault (Hymenoptera: Eulophidae); and the predators; Campylomma austrina Malipatil and Campylomma sp (Hemiptera: Miridae), Geocoris lubra Kirkaldy and Stigmatonotu minutum Malipatil (Hemiptera: Lygaeidae), Orius armatus Gross and Orius tantillus (Motschulsky) (Hemiptera: Anthocoridae), Haplothrips victoriensis Bagnall (Thysanoptera: Phlaeothripae), ? Amblyseius sp (Acarina: Phytoseiidae); and spiders of the families Aranidae, Oxyopidae, Salticidae, Theridiidae, Thomisidae and Uloboridae (NT DPIF records; Layland and Brown pers. comm.; unpublished data). T. palmi has been shown to be tolerant to a wide range of insecticides (Suzuki et al. 1982; De Bon and Rhino 1989; Cooper 1990; Kawai 1990; Layland and Brown pers. comm.). Layland and Brown (pers. comm.) working from 1989 to 1992 found that prothiophos and promecarb gave the best control of the thrips, but prothiophos had a long with-holding period and consequently was unlikely to be registered for vegetables. By 1995, promecarb had been withdrawn from sale and endosulfan was widely used for control of T. palmi on vegetables (unpublished data). Endosulfan gave poor control and it was decided to test the effect of soft chemicals, such as soaps and petroleum spray oils, on melon thrips. Bioassays showed that potassium soap (Natrasoap®) was the most active and at 16 ml/L gave up to 87% and 67% mortality of T. palmi larvae and adults respectively (unpublished data). Eggplant is planted as transplanted seedlings at the start of the Top End dry season and grown under trickle irrigation (Young and Zhang 1998). The period from transplanting to flowering is about 25 days. The crop is often ratooned at six to 12 monthly intervals extending the life of the crop to 18 months. During the dry season of 1995 a grower from Knuckey's Lagoon reported severe stunting of eggplant seedlings as a result of attack by T. palmi. There were two adjacent eggplant crops at Knuckey's Lagoon, one ratooned in July 1995 and the other transplanted as seedlings (plant crop) in early August. Each crop consisted of five, 80-metre rows. When the crops were inspected on 15 September 1995, the leaves of both crops were heavily infested with T. palmi. The ratoon crop had started to flower but the plant crop was badly stunted. 102 It was decided to try to control the pest using potassium soap (Natrasoap) and to compare the soap treatment to conventional chemical insecticides. Additionally, it was decided to look at the phenology and natural enemies of T. palmi on eggplant. Materials and Methods Effect of potassium soap, Natrasoap®, on T. palmi populations Ten leaf samples were taken at random from the stunted plant crop every week, for 11 weeks, starting on 15 September 1995 (week one). The leaf samples were returned to the laboratory and counts made of the numbers T. palmi adults and larvae. The crop was sprayed with potassium soap at 16 ml/L using a knapsack sprayer, three days after sampling during weeks two, three and four. Phenology and natural enemies of T. palmi on eggplant From 20 November 1995, weekly leaf samples were taken from the plant and ratoon crops, until 12 July 1996 for the plant crop and 22 July for the ratoon crop. Each week three rows were selected from each crop and ten leaves sampled at random from each row, making a total of 30 leaves per crop. Counts were made of the numbers of adults and larvae T. palmi per leaf. Predators found on the leaves were tested in the laboratory to ascertain if they would feed on T. palmi. Up to the time of the first sample the plant crop had only been sprayed with potassium soap. The ratoon crop had been sprayed with both permethrin and endosulfan at fortnightly intervals from early October until 20 November 1995. Effect of potassium soap on T. palmi and its natural enemies. The plant crop was ratooned on 15 July 1996. Sampling of leaves started on the 20 August (week one) and continued until the 1 November 1996 (week 11). The crop was sprayed with potassium soap, at 16 ml/L, after sampling on weeks two, three, four and eight. There was one spray of endosulfan, during week seven, to control the eggfruit caterpillar, Sceliodes cordalis (Doubleday) (Lepioptera: Pyralidae). Ten leaves were sampled at random from each row per week, making a total of 50 leaves. Counts were made of the number of T. palmi adults and larvae as well as thrips predators. 103 Results Effect of potassium soap on T. palmi populations Sprays of potassium soap reduced populations of T. palmi. Counts at week one recorded means of 14.1 adults and 227.4 larvae per leaf and at week two, 29.8 adults and 211.3 larvae per leaf. The population, of adult and larvae T.palmi, was progressively reduced after each application of potassium soap to 0.6 adults and 6.5 larvae per leaf at week eight. While the numbers of larvae increased at week nine to 30.5 per leaf, at week 11 mean numbers of adults and larvae were 3.1 and 5.2 per leaf (Figure 1). By week seven the crop had recovered from the thrips attack and normal growth had resumed. Mean no T. palmi adults & larvae/leaf 240 Larvae Adults 180 120 60 0 1 2 3 4 5 6 7 8 9 10 11 Weeks Fig. 1 Mean number of T. palmi adults & larvae per leaf Phenology and natural enemies of T. palmi The wet season started during November and finished in mid April (Table 2). Table 2. Rainfall in mm over the period of the study (Bureau of Meteorology, NT Regional Office, site 014015, Darwin) Oct 59.4 1995 Nov 291.8 Dec 198.4 Jan 292.4 Feb 274.0 Mar 272.0 1996 Apr May 137.4 0.0 Jun 0.0 Jul 0.0 Aug 1.0 Mean numbers of T. palmi on the plant crop were 1.4 and 7.0 per leaf for adults and larvae, respectively, at the start of sampling (Figure 2 and 3). The numbers declined to less than one per leaf by the end of February 96 and remained at these levels until June/ July 96, when the population began to rise again. The last sample on 12 July 1996 recorded means of 2.5 and 16.5 per leaf for adults and larvae. Mean numbers of T. palmi on the ratoon crop were 11.6 and 53.2 per leaf for adults and larvae, respectively, at the start of sampling (Figures 4 and 5). The population declined sharply during January 1996 to 0.2 and 0.4 adults and larvae on 29 January. Numbers remained at low levels 104 until June 1996, when the population began to rise. The last sample on 22 July 1996 recorded means of 8.7 and 95.8 per leaf for adults and larvae. Predators recovered during sampling were: larvae and adults of predatory thrips, H. victoriensis; nymphs and adults of Deraeocoris sp. (Hemiptera: Miridae); larvae of Mallada basalis (Walker) (Neuroptera: Chrysopidae); ? Amblyseius sp. and adult Orius sp. All the predators fed on T. palmi larvae in the laboratory. M. basalis and Deraeocoris sp. also fed on Aphis gossypii Glover (Hemiptera: Aphididae) and Tetranychus urticae Koch (Acarina: Tetranychidae). Deraeocoris sp. and Orius sp. adults were observed feeding on up to seven T. palmi larvae per day. Adults and nymphs of Deraeocoris sp. appeared able to survive solely on T. palmi larvae. H. victoriensis adults fed on up to two T. palmi larvae per day. No pathogens or parasites were recovered during sampling. 105 Mean no. of T. palmi larvae per leaf 24 20 16 8 12 Mean+SE Mean-SE Mean 20/11/95 27/11/95 04/12/95 11/12/95 18/12/95 27/12/95 02/01/96 08/01/96 15/01/96 23/01/96 29/01/96 06/02/96 12/02/96 19/02/96 26/02/96 04/03/96 12/03/96 18/03/96 25/03/96 01/04/96 09/04/96 15/04/96 22/04/96 29/04/96 10/05/96 17/05/96 24/05/96 31/05/96 07/06/96 14/06/96 21/06/96 05/07/96 12/07/96 22/07/96 4 0 4 3 2 on the plant crop (Potassium soap) Fig. 2 & 3 Mean numbers of T. palmi adults & larvae per leaf 20/11/95 27/11/95 04/12/95 11/12/95 18/12/95 27/12/95 02/01/96 08/01/96 15/01/96 23/01/96 29/01/96 06/02/96 12/02/96 19/02/96 26/02/96 04/03/96 12/03/96 18/03/96 25/03/96 01/04/96 09/04/96 15/04/96 22/04/96 29/04/96 10/05/96 17/05/96 24/05/96 31/05/96 07/06/96 14/06/96 21/06/96 05/07/96 12/07/96 22/07/96 1 0 120 100 80 60 40 0 16 14 12 10 8 6 4 2 0 on the ratoon crop (endosulfan & permethrin) Fig. 4 & 5 Mean numbers of T. palmi adults and larvae per leaf 20/11/95 27/11/95 04/12/95 11/12/95 18/12/95 27/12/95 02/01/96 08/01/96 15/01/96 23/01/96 29/01/96 06/02/96 12/02/96 19/02/96 26/02/96 04/03/96 12/03/96 18/03/96 25/03/96 01/04/96 09/04/96 15/04/96 22/04/96 29/04/96 10/05/96 17/05/96 24/05/96 31/05/96 07/06/96 14/06/96 21/06/96 05/07/96 12/07/96 22/07/96 20 106 20/11/95 27/11/95 04/12/95 11/12/95 18/12/95 27/12/95 02/01/96 08/01/96 15/01/96 23/01/96 29/01/96 06/02/96 12/02/96 19/02/96 26/02/96 04/03/96 12/03/96 18/03/96 25/03/96 01/04/96 09/04/96 15/04/96 22/04/96 29/04/96 10/05/96 17/05/96 24/05/96 31/05/96 07/06/96 14/06/96 21/06/96 05/07/96 12/07/96 22/07/96 Mean no. of T. palmi adults per leaf Mean no. of T. palmi larvae per leaf Mean no. of T. palmi adults per leaf Effect of potassium soap on T. palmi and its natural enemies. Initially the number of T. palmi adults increased from weeks one to two, from 159 to 254 per 50 leaves, then decreased in weeks three to six from 103 to 10 per 50 leaves (Figure 6). Adult numbers again increased in weeks seven to nine, from 56 to 98, before declining to three per 50 leaves at week 11. Mean counts of T. palmi larvae per half leaf increased in weeks one to three, from 28.6 to 38.2, then decreased in weeks four to seven, from 13.7 to 2.0, before increasing to 19.3 in week 8 and then declining to 1.3 in week 11 (Figure 7). No. adult T. plami / 50 leaves Predators recovered in order of abundance were; H. victoriensis, Deraeocoris sp., M. basalis larvae and Orius sp. (Figures 8, 9 and 10). H. victoriensis was most numerous in weeks one to six, with numbers ranging from 1 to 14 per 50 leaves, while Deraeocoris sp. and M. basalis were most common from weeks 6 to 11. M. basalis numbers plateaued during weeks eight to eleven. Orius sp was only observed in weeks two and 11. The single endosulfan application in week seven did not appear to affect H. victoriensis, Deraeocoris sp. or M. basalis. 280 240 200 160 120 80 40 0 1 2 3 4 5 6 7 8 9 10 11 10 11 Week Mean no. of T. palmi larvae / half leaf Fig. 6 No. T. palmi adults on eggplant sprayed with potassiom soap 50 40 Mean+SE Mean-SE 30 Mean 20 10 0 1 2 3 4 5 6 7 8 9 Fig. 7 Mean number of T. palmi larvae per half leaf on eggplant sprayed with potassium soap 107 16 14 T. palmi larvae H. victoriensis 12 30 10 8 20 6 4 10 No. H.victoriensis/50 leaves Mean no. T. palmi larvae / half leaf 40 2 0 0 1 2 3 4 5 6 7 8 9 10 11 WEEK Fig. 8 No. of T. palmi larvae and H.victoriensis 4 T. palmilarvae (L) M.basali (R) No. of M. basalis larvae/50 leaves Mean no. of T. palmi larvae / half leaf 40 3 30 2 20 1 10 0 0 1 2 3 4 5 6 7 8 9 10 11 WEEK 40 4 30 3 T. palmi larvae(L) Deraeocosp. (R) 20 2 10 1 0 0 1 2 3 4 5 6 7 8 9 WEEK Fig. 10 Number of T. palmi larvae & Deraeocoris sp. 108 10 11 No. of Deraeocoris sp./50 leaves Mean no. of T. palmi larvae / half leaf Fig. 9 Number of T. palmi larvae & M. basalis larvae Discussion Effect of potassium soap on T. palmi populations The initial field trial showed that potassium soap rapidly reduced the population of T. palmi larvae from a mean of 211.3 per leaf at week two to 8.3 at week five (Fig.1). This result confirmed previous bioassay data on the efficacy of potassium soap against T. palmi. Phenology and natural enemies of T. palmi The population of T.palmi, on the plant crop was low from the start of sampling and declined further as the wet season progressed (Figures 2 and 3). On the ratoon crop the population was initially high, probably as a result of permethrin and endosulfan killing T. palmi predators (Figures 4 and 5). The T. palmi population on the plant crop was probably controlled by natural enemies from the start of sampling to mid January 96 (Figures 2 and 3). By mid January the ratoon crop population had dropped to similar levels to that of the plant crop (Figures 4 and 5). The populations on both crops started to increase again during June about six weeks after the end of the wet season. These population trends confirm earlier observations on the high mortality of prepupae and pupae in saturated soil. Anthocorid and mirid predators would seem to offer great potential for the control of T. palmi in the NT. Hirose (1990) also recorded anthocorid and mirid predators in Thailand. Walker (1994) listed nine species of anthrocorid, particularly Orius spp., considered to play an important role in the biological control of T. palmi in Japan, China and South East Asia. However, experience with T. palmi on eggplant suggests that anthocorids are not as important as predators in the NT as they are in Asia (unpublished data). The larval parasite Ceranisus menes (Walker) (Hymenoptera: Eulophidae) is regarded as the most important natural enemy of T. palmi in Thailand, but the parasite has not been recorded from T. palmi in the NT (Hirose 1990; Hirose et al 1993). The effect of potassium soap on T. palmi and its natural enemies The mean counts of T. palmi larvae per half leaf and the count of adult T. palmi adults per 50 leaves followed approximately the same population trends (Fig. 6 and 7). While potassium soap reduced the population of T. palmi to relative low numbers, the low counts of predators makes it difficult to assess their influence on the prey population (Figures 8, 9 and 10). Potassium soap did not appear to affect the three main predators. The brief resurgence of T. palmi during week 8 may have been the result of the single spray of endosulfan during week seven causing predator mortality, although there does not appear to be any discernible effect on H. victoriensis, M. basalis and Deraeocoris sp. (Figures 8, 9 snd 10). The counts of H. victoriensis and Deraeocoris sp. approximately follow the population trend of T. palmi larvae with a one to two week time lag (Figures 8 and 10). While both species have been recorded feeding on arthropod species other than thrips it would appear from the results that they are primarily thrips predators (Bailey and Caon 1986). Observations on Deraeocoris sp. since the current work have shown that the predator is usually detected just as the population of T. palmi begins to decline (unpublished data). M. basalis counts plateauxed after week eight at a time when the T. palmi population had declined (Figure 9). This would suggest that M. basalis does not feed primarily on thrips. Later feeding trials revealed that M. basalis showed no preference between T. palmi, T. urticae or A. gossypii (unpublished data). A more intensive sampling program would be necessary to accurately establish a relationship between predator numbers and T. palmi populations. There is a need to repeat the feeding trials with the three main predators to accurately determine their influence on T. palmi populations. The results of the three trials show that T. palmi can be controlled initially by using potassium soap, allowing predators to control the remaining population. It is reasonable to assume from the results that T. palmi is largely an insecticide induced pest in the NT. 109 Acknowledgments Thanks are due to Mr G. Padovan, Knuckey's Lagoon, for allowing us to use his crops; Dr M. Malipatil, Institute of Horticultural Development, Agriculture Victoria, for identification of Deraeocoris sp. and Deanna Chin, Entomology Branch, NT DPIF, for comments on the manuscript. References Bailey, P. and Caon, G.(1986). Predation on two spotted mite, Tetanychus urticae Koch (Acarina: Tetranychidae) by Haplothrips victoriensis Bagnall (Thysanoptera: Phlaeothripidae) and Stethorus nigripes Kapur (Coleoptera: Coccinellidae) on seed lucerne crops in South Australia. Australian Journal of Zoology 34: 515-525. Bournier, J.P. (1983). A polyphagous insect: Thrips palmi Karny important cotton pest in the Philippines. Cotton et Fibres, Tropicale 38: 286-289. Cooper, B. (1991). Status of Thrips palmi Karny in Trinidad. FAO Plant Protection Bulletin 39: 45-46. De Bon, H. and Rhino, B. (1989). Lutte contre Thrips palmi Karny a la Martinque. L' Agronomie Tropicale 44-2: 129-136. Hirose, Y. (1990). Prospective use of natural enemies to control Thrips palmi (Thysanoptera:Thripidae). In FFTC-NARC International seminar on 'The use of natural enemies to control agricultural pests'. FFTC book No. 40: 135-141. Tukubu-gun, Japan. Hirose, Y. Katita, H., Takagi, M., Okajima. S., Napompeth, B. and Buranapanichan, S. (1993). Natural enemies of Thrips palmi and their ffectiveness in the native habitat, Thailand. Biological Control 3: 1-5. Houston, K. J., Mound, L. A., and Palmer, J. M. (1991). Two pest thrips (Thysanoptera) new to Australia, with notes on the distribution and structural variation of other species. Journal of the Australian Entomological Society 30: 231-232. Johnson, M. W. (1986). Population trends of a newly introduced species, Thrips palmi (Thysanoptera: Thripidae), on commercial watermelon plantings in Hawaii. Journal of Economic Entomology 79: 718-720. Kajita, H., Hirose, Y., Takagi, M., Okajima, S., Napompeth, B. and Buranapanichan, S. (1996). Host plants and abundance of Thrips palmi Karny (Thysanoptera: thripidae), an important pest of vegetables in South East Asia. Applied Entomology and Zoology 31: 87-94. Kawai, A. (1990). Life cycle and population dynamics of Thrips palmi Karny. Japan Agricultural Research Quarterly 23: 282-288. Kawai, A. (1990). Control of Thrips palmi Karny in Japan. Japan Agricultural Research Quarterly 24: 43-48. Layland, J. (1991). Thrips palmi Agnote No. 481. NT Department of Primary Industry and Fisheries. Darwin NT, Australia. Layland J. K., Upton, M. and Brown, H. H. (1994). Monitoring and identification of Thrips palmi Karny (Thysanoptera: Thripidae). Journal of the Australian Entomological Society 33: 169-173. Suzuki, H., Tamaki, S. and Miyara, A. (1982). Physical control of Thrips palmi Karny. Proceedings of the Association of Plant Protection of Kyushu. 28: 134-137. 110 Tsai, J. H., Yue, B., Webb, S. E., Funderburk, J. E. and Hsu, H. T. (1995). Effects of host plant and temperature on growth and reproduction of Thrips palmi (Thysanoptera: Thripidae). Environmental Entomology 24: 1598-1603. Walker, A. K. (1994). A review of the pest status and natural enemies of Thrips palmi. Biocontrol News and Information 15: 7N-10N. Waterhouse, D. F. and Norris, K. R. (1987). Thrips palmi in Biological Control Pacific Prospects. Australian Centre for International Agricultural Research, p 90-94. Inkata Press, Melbourne, Australia. Young, G. R. and Zhang, L. (1998). IPM of snake bean, Vigna unguiculata spp. sesquipedalis, in the top end of the Northern Territory. In Young Ed. 6th Workshop on Tropical Agricultural Entomology. Darwin, NT. 11th to 15th of May 1998. NT Department of Primary Industries and Fisheries, Darwin NT. 111 FACTORS INFLUENCING THE SPATIAL DISTRIBUTION OF THE TEA MOSQUITO BUG, HELOPELTIS PERNICIALIS, IN CASHEW PLANTATIONS Renkang Peng, Keith Christian and Karen Gibb Faculty of Science Northern Territory University Darwin NT 0909 Abstract The tea mosquito bug, Helopeltis pernicialis, is the most important insect pest in cashew plantations in the Northern Territory. In order to provide plantation managers with suitable methods of monitoring and control of the pest, the factors that are responsible for spatial distributions of H. pernicialis in cashew orchards were investigated using rearing, field surveys and observations in 1993 and 1994 at Wildman River Plantation, Northern Territory. The vertical distribution of flushing shoots damaged by H. pernicialis suggested that this insect pest was more active at the bottom level than at the middle or top levels of tree canopies. Based on field observations, H. pernicialis were more active in the early morning and late afternoon than the midday. Eggs of this insect pest were only found in foliar and floral flushing shoots. In these shoots, over 75% of eggs were found at the base of young leaf petioles, the tender stems between the young leaves and the tender stems of flower panicles. Bag rearing showed that 85% of H. pernicialis eggs hatched from shoots which were cultured with water, but no eggs hatched from shoots which were kept under dry conditions. Bag rearing also showed that both nymphae and adults preferred feeding on the tender parts of flushing shoots. Therefore, it is suggested that flushing shoots are essential for H. pernicialis to survive. Spatial distribution of H. pernicialis was patchier in non-sprayed orchards than in sprayed orchards. The distribution in the sprayed orchards was associated with variability in the timing of flushing among cashew varieties, while the distribution in the non-sprayed orchards was attributed to the distribution of a predator, the green ant. The use of green ants as a biological control agent in cashew plantations is discussed. Introduction Helopeltis spp are the most serious insect pests of cashews in the tropics (Stonedahl 1991). Helopeltis pernicialis (Stonedahl, Malipatil and Houston) is the most important pest in cashew growing areas in the Northern Territory, such as Wildman River Plantation, Costal Plains Station, Howard Springs and Katherine (Houston and Malipatil 1991). This insect causes serious damage to the tender shoots, the inflorescence, the developing apple and the young nut. Crops can be completely destroyed by this pest (Peng et al. 1997a). Basic bioecology of H. pernicialis was described by Houston and Malipatil (1991) and Stonedahl et al. (1995), but the factors that are associated with the spatial distribution of this pest in cashew orchards were not investigated. This information is essential for the monitoring and control of H. pernicialis. The aim of this study was to determine the factors that influence the distribution of this pest in cashew orchards. Materials and methods Field and laboratory rearing, surveys and observations were done in 1993 and 1994 at Wildman River Plantation (12o64'S 131o87'E), which is in the wet-dry tropics of the Northern Territory of Australia. The trees used for this study were seven years old. A flexible netting bag (45 × 28 × 18 cm) was developed to confine insects to an area on a field cashew tree (Peng et al. 1998). Microclimate and flushing shoot conditions inside the bag are closer to the field conditions than inside conventional insect cages. When nutrition of flushing shoots inside the bag is poor due to insect feeding, the bag can be easily transferred to other flushing shoots. This system was used in the field to rear H. pernicialis and to determine the changes in damage symptoms. In order to determine whether fresh flushing shoots are 113 essential for development of H. pernicialis eggs, twenty healthy flushing shoots with newly-laid eggs were bagged; ten of them were cultured with water to keep the shoots fresh and the other ten were kept under dry conditions. The numbers of hatched eggs were counted after two weeks. Previous observations indicated that H. pernicialis tend to hide when they are disturbed, but their damage symptoms are easily recognised. Therefore, to determine the vertical distribution of H. pernicialis within a tree, flushing shoots with fresh damage symptoms were used as indicators. Each tree canopy was divided to nine zones; two zones at the top level, three at the middle level and four at the bottom level. In each zone, the total number of flushing shoots and the number of shoots freshly damaged by H. pernicialis were recorded. Damage that was one to two days old was regarded as fresh damage (Table 1). Every cashew tree in a block of 1.3 ha was inspected. To determine the oviposition sites of H. pernicialis on cashew trees, foliar and floral flushing shoots and non-flushing shoots at different positions of cashew trees were removed and taken to the laboratory for examination under a binocular microscope. H. pernicialis eggs were oviposited into the cashew tissue with two filaments of unequal length protruding from the surface of the tissue, and they were easily seen under a dissecting microscope. For the spatial distribution of H. pernicialis in the field, two types of cashew orchards were used, a non-sprayed block of 2.4 ha and a sprayed block of 6 ha. Every tree in the non-sprayed block was checked, and in the sprayed block every other tree in a row and every other row of trees were checked. A total of 410 trees in the non-sprayed block and a total of 235 trees in the sprayed block were inspected. Four field surveys were completed during the outbreak of this insect pest over two years (August 1993 and August 1994 in the sprayed block; November 1993 and September 1994 in the non-sprayed block). Preliminary results showed that the green ant, Oecophylla smaragdina (F.), was the most abundant predator in cashew orchards (Peng, et al. 1994). For each tree, the number of shoots freshly damaged by H. pernicialis, the total number of flushing shoots and the dominant ant species in the tree canopy were recorded. The ratio of the variance (s2) to the mean (M) was used to measure the distributions of H. pernicialis in the field (Taylor 1984). If s2/M < 1, = 1 or > 1, the distribution is even, random or aggregated respectively. One-way ANOVA was used to compare H. pernicialis damage among levels of the tree canopy, the number of eggs oviposited in different parts of foliar and floral flushing shoots and the damage among cashew varieties. The percentage data were transformed before one-way ANOVA was applied. Results Table 1 shows that flushing shoots sucked by H. pernicialis were identifiable by bruised marks after one day and the marks were darker with light-brown colour after two days. The marks turned to dark brown oval-shapes after three days, and the oval-shaped areas were reddish black after four days. From bag rearing, it was observed that the fresh damage symptoms only occurred on young leaves, tender stems and tender flower stems. Laboratory experiments showed that (85 ± 11)% of H. pernicialis eggs hatched from the shoots that were cultured with water, but no eggs hatched from the shoots that were kept under dry conditions. 114 Table 1. Changes in damaged area by Helopeltis pernicialis, n = 20 shoots Sucking from tender parts of flushing shoots after 1 day 2 days 3 days 4 days Bruised marks appeared in the damaged area The bruised marks expanded and became darker (light brown) The bruised marks became oval-shaped and dark brown in colour The oval-shaped areas turned into reddish black patches Table 2. Distribution fresh damage (%) of H. pernicialis within a cashew tree canopy, n = 235 trees Level Top Middle Bottom 4 Mean damage at each level 51.2 ± 39.7 22.4 ± 26.3 a 35.8 ± 34.6 b 47.9 ± 38.7 Mean damage (%) in zones ± SD 1 2 3 21.1 ± 27.2 34.2 ± 34.1 45.3 ± 38.2 23.6 ± 27.2 38.1 ± 35.3 47.1 ± 38.3 35.2 ± 34.4 48.1 ± 38.5 Means for each level followed by the same letter are not significantly different at the 5% level. The percentages of freshly damaged flushing shoots at different levels of tree canopies are shown in Table 2. Significantly more flushing shoots were damaged by H. pernicialis at the bottom level than at the middle or top levels. No eggs were found in non-flushing shoots. The numbers of eggs that were found in foliar and floral flushing shoots are shown in Tables 3 and 4. In the foliar flushing shoots, the base of young leaf petioles and the tender stems between the young leaves contained 75% of the total eggs, which was significantly higher than other parts of the shoots (Table 3). In the floral shoots, the base of tender flower stems and the tender flower stems had significantly higher numbers of eggs (78%) than any other parts of the inflorescence (Table 4). Table 3. Distribution of H. pernicialis eggs on foliar flushing shoots (n = 24 shoots) Position Numbers of eggs Total Mean ± SD Bud Young leaf mid-rib Young leaf petiole Base of young leaf petiole Tender stem between young leaves Mature leaf mid-rib Mature leaf petiole Mature stem (green) 0.0 0.46 ± 0.76 0.42 ± 0.76 4.00 ± 4.45 5.28 ± 6.12 0.96 ± 1.95 1.33 ± 2.59 0.0 d c c a a b b d 0 11 10 96 132 23 32 0 Means for each position followed by the same letter are not significantly different at the 5% level. 115 Table 4. Distribution of H. pernicialis eggs on floral flushing shoots (n = 31 panicles) Position Numbers of eggs Total Mean ± SD Flower bud Tender flower stem Base of tender flower stem Base of fruit Mature leaf petiole Base of mature leaf petiole c a a c b b 0.10 ± 0.30 1.84 ± 1.25 1.58 ± 1.60 0.10 ± 0.51 0.35 ± 1.09 0.42 ± 0.75 3 57 49 3 11 13 Means for each position followed by the same letter are not significantly different at the 5% level. Spatial distributions of H. pernicialis varied with the types of orchards (Table 5). The ratios of the variance to the mean were all much larger than 1, and the ratios were even larger in the nonsprayed area than in the sprayed area. This result was consistent between the two years (Table 5). An analysis in the sprayed area, where green ants were absent, suggested that there was no significant difference of Helopeltis damage between the trees occupied by different ant species and the trees without ants. Helopeltis damage in relation to different cashew varieties was then investigated, and the results are shown in Table 6. Thirteen varieties were all damaged, but A1, K10-2 and V30/3, which flushed early, had more damage than the other varieties. The variety V13 and V10, which flushed late, were less damaged than the other varieties (Table 6). In the non-sprayed area, green ants were abundant and occupied 57% of the trees in 1993 and 70% in 1994 (Table 7; Peng et al. 1997b). Table 5. Spatial distributions of H. pernicialis presented as percentage fresh damage in different types of cashew orchards Month/Year Type of orchard 8/1993 11/1993 8/1994 9/1994 Sprayed Non-sprayed Sprayed Non-sprayed 116 Mean (M) 37.78 2.74 48.76 4.81 Variance (s2) 492.84 53.88 913.85 125.22 s2/M 13.04 19.66 18.74 26.03 Sample size (trees) 235 409 235 410 Table 6. The average fresh damage of H. pernicialis on different varieties of cashews which were evenly planted in one block Variety Mean percentage damage ± SEM A1 A2 ALA273-1 H3-17 K10-2 NDR2-1 V10 V13 V30/3 V59/2 V77 VLA3 VMR/1 74.4 ± 4.4 a 39.1 ± 6.2 b 30.6 ± 8.2 b 25.8 ± 6.2 b 53.3 ± 8.3 ab 40.6 ± 6.2 b 22.9 ± 5.0 b 18.7 ± 4.9 bc 53.1 ± 7.2 ab 25.4 ± 7.3 b 35.2 ± 7.5 b 36.5 ± 8.2 b 36.6 ± 8.8 b Sample size (trees) 18 17 19 18 18 18 19 18 18 18 18 17 18 Flushing period Early Middle Middle Late Middle Middle Late Late Early Middle Middle Middle Middle Means for each variety followed by the same letter are not significantly different at the 5% level. When cashew trees were grouped according to the trees occupied by different species of ants and the trees without ants (Table 7), it can be seen that trees occupied by O. smaragdina, Iridomyrmex sanquineus and Opisthopsis haddoni had significantly less damage than trees occupied by Paratrechina sp or trees without ants. A similar result was also obtained in 1994 field survey (Peng et al. 1998). Discussion Shelter appears to be an important factor in determining the vertical distribution of the pest. Significantly more flushing shoots were damaged at the bottom level than the middle or top levels of tree canopies (Table 2). The major difference was that flushing shoots at the bottom level were more sheltered than those at middle and top levels. Table 7. The average fresh damage caused by H. pernicialis in relation to trees occupied by different species of ants Trees with Oecophylla smaragdina Iridomyrmex sanguineus Opisthopsis haddoni No ants O. haddoni and Paratrechina sp Paratrechina sp Number of trees Mean percentage damage ± SEM 234 8 22 94 12 0.7 ± 0.2 a 1.9 ± 1.4 a 0.4 ± 0.3 a 5.4 ± 0.9 b 7.0 ± 2.6 b 37 9.6 ± 2.5 c Means for each species followed by the same letter are not significantly different at the 5% level. Houston and Malipatil (1991) showed that H. pernicialis were more abundant on mature cashew trees than on young trees. The main difference between mature and young trees was that there were many gaps between young trees, but almost no gaps between mature trees. Field observations in this study showed that H. pernicialis were more active in the early morning and late afternoon than the midday. This activity pattern suggests that this insect avoids strong light and high temperature. 117 Tender tissue of flushing shoots is essential for H. pernicialis. From the results of bag rearing, this insect preferred to feed on young leaves, tender stems and tender flower stems more than any other parts of flushing shoots. For oviposition, this insect pest preferred the base of young leaf petioles and the tender stems between young leaves more than other parts of the shoots in foliar flushing shoots (Table 3). In floral flushing shoots, a larger proportion of eggs was found in the base of tender flower stems and the tender flower stems than elsewhere (Table 4). From the fact that H. pernicialis eggs only hatched from the shoots, which were cultured with water, it is suggested that living flushing shoots are also essential in the development of the eggs. Because of this behaviour, plantation managers can detect the early presence of this insect pest by examining the tender parts of flushing shoots at the bottom level of tree canopies. Such examinations are very important for controlling H. pernicialis. Houston and Malipatil (1991) reported that H. pernicialis distribution in cashew orchards was patchy. This has been confirmed by the data obtained in the sprayed block in this study (s2/M = 13 in 1993 and 18.7 in 1994, Table 5). In this block, H. pernicialis fed on those trees which had abundant flushing shoots, such as varieties A1, K10-2 and V30/3 which flushed earlier than the other varieties (Table 6). Varieties V13 and V10 were damaged less by this pest than other varieties (Table 6) because these varieties flushed late, which did not coincide with the early incidence of the pest. By the time of this survey, more adults of this pest were found on these two varieties than nymphae, which suggested that they flew in from other varieties for feeding and oviposition. Therefore, it is concluded that the patchy distribution of Helopeltis in the sprayed block was associated with variability in the timing of flushing among cashew varieties. None of the cashew varieties tested was resistant to H. pernicialis damage. A very patchy distribution was found in the non-sprayed block (s2/M = 19.7 in 1993 and 26 in 1994, Table 5). In this area, a high proportion of trees occupied by O. smaragdina was almost free from H. pernicialis damage (Table 7), and O. smaragdina distribution was patchy (Peng et al. 1997b). However, trees without ants or with Paratrechina sp. were substantially damaged by the pest (Table 7). Therefore, the very patchy distribution of the pest in the non-sprayed block is closely associated with the distribution of O. smaragdina. O. smaragdina have two forms of food: sugar and meat. They can obtain sufficient sugar from a large number of extrafloral nectaries on the tender leaves, inflorescences, flowers and developing nuts (Rickson and Rickson 1998). The meat the ants rely on in cashew orchards includes the insect pests that attack flushing shoots, inflorescences, flowers and developing nuts (Peng et al. 1995, 1997a,b). In order to catch their prey and to gather nectar, O. smaragdina continuously patrol flushing terminals, inflorescences and developing nuts. This activity greatly reduces insect pest numbers and prevents insect pests from feeding and ovipositing on the trees. In the case of H. pernicialis, O. smaragdina is particularly effective because they not only feed on Helopeltis nymphae, but also repel Helopeltis adults from feeding on or ovipositing in flushing shoots. It is concluded that O. smaragdina can be used as biological control agents to control the tea mosquito bug in cashew orchards. Acknowledgements The research was supported by the Northern Territory University and the Rural Industries Research and Development Corporation. We thank Wildman River Plantation and Mr Ian Duncan for providing us with study sites and necessary facilities. We also thank Mr Craig Palmer and Mrs Keeley Palmer for their help with field work. 118 References Houston, W. and Malipatil, M. (1991). Bioecology of cashew insects at Wildman River, Northern Territory. 58 pp. A report to the Rural Industries Research and Development Corporation, Canberra. DPIF, Darwin. Peng, R.K., Christian, K. and Gibb, K. (1994). The effect of the green ant on cashew insect pests with particular reference to the tea mosquito bug (Helopeltis sp.). Pp 75-81. Seventh Cashew Research and Development Workshop, May 17-19, 1994, Cairns, Australia. Peng, R.K., Christian, K. and Gibb, K. (1995). The effect of the green ant, Oecophylla smaragdina (Hymenoptera: Formicidae), on insect pests of cashew trees in Australia. Bull. Entomol. Res. 85: 279-284. Peng, R.K., Christian, K. and Gibb, K. (1997a). Control threshold analysis for the tea mosquito bug, Helopeltis pernicialis (Hemiptera: Miridae) and preliminary results of the control efficiency by the green ant, Oecophylla smaragdina (Hymenoptera: Formicidae) in northern Australia. Int J. Pest Man. 43: 233-237. Peng, R.K., Christian, K. and Gibb, K. (1997b). Distribution of the green ant, Oecophylla smaragdina (F.) (Hymenoptera: Formicidae), in relation to native vegetation and the insect pests in cashew plantations in Australia. Int. J. Pest Man. 43: 203-211. Peng, R.K., Christian, K. and Gibb, K. (1998). Impact of native vegetation on cashew insect pests with particular reference to the most important pest - Helopeltis pernicialis. 70 pp, Rural Industries Research and Development Corporation, Canberra, Australia. Rickson, F.R. and Rickson, M.M. (1998). The cashew nut, Anacardium occidentale (Anacardiaceae), and its perennial association with ants: extrafloral nectary location and the potential for ant defense. Am. J. Bot. 85: 835-849. Stonedahl, G.M. (1991). The oriental species of Helopeltis (Heteroptera: Miridae): a review of economic literature and guide to identification. Bull. Entomol. Res. 81: 465-490. Stonedahl, G.M., Malipatil, M.B. and Houston, W. (1995). A new mirid (Heteroptera) pest of cashew in Northern Australia. Bull. Entomol. Res. 85: 275-278. Taylor, L.R. (1984) Assessing and interpreting the spatial distributions of insect populations. Ann. Rev. Ent. 29: 321-57. 119 THE PUZZLE OF INSECT DAMAGE IN CONVENTIONALLY CULTIVATED AND NO TILLAGE PLOTS AT THE DOUGLAS DALY RESEARCH FARM G.R. Young and E.S.C. Smith Entomology Branch Department of Primary Industry and Fisheries GPO Box 990 Darwin NT 0801 Abstract A conventional cultivation/direct drill trial has been conducted for 13 years at Douglas Daly Research Farm, Northern Territory. The trial was carried out on a red lateritic soil. A variety of legume and poaceaous crops have been planted each wet season. Surveys of crops, one week post emergent, showed that contrary to expectations the direct drill plots sustained more insect damage than the conventionally cultivated plots. Pitfall trapping was carried out in till and no-till plots, from August to the start of the wet season, during the years 1995 to 1998. The object was to look at any differences in invertebrate fauna between till and no till plots to try to explain the level of insect damage in the till plots. It seems that minimum cultivation will not reduce insect damage under the soils and arid tropical conditions existing in the Top End of the Northern Territory. Key words: conventional cultivation, direct drill, phytophagus insects, predatory ants. Introduction The term conservation tillage covers both reduced and no-tillage (direct drill) practices. Conservation tillage has been developed to protect soil from water and wind erosion (Stinner and House 1990). Additionally, conservation tillage reduces the rate of organic matter breakdown and the decline of soil structure (Cogle et al. 1991). Sustainable cropping systems are needed in the arid tropics in situations where fragile soils and periodic high intensity rainfall leads to soil erosion (Robertson et al. 1994). Stinner and House (1990) state that tillage influences the invertebrate fauna in three ways: physical disturbance, residue placement and effects on weed communities. With no tillage the crop residues are left on the soil surface. The organic matter is slowly incorporated into the soil by invertebrate activity. The litter layer tends to stabilise soil temperatures and moisture levels providing a favourable habitat for soil invertebrates. It is generally thought that retention of crop residues on the soil surface favours decomposer and predatory soil fauna (Robertson 1994). A surface layer of organic litter attracts decomposers which in turn supports a diverse and abundant predatory fauna (Robertson 1993). However, the influence of invertebrate pests on crop production in no tillage systems remains a controversial issue, with the effects of pests varying among soils, crops and prevailing climate (Lal 1986). A conventional cultivation/direct drill trial has been conducted at the Douglas Daly Research Farm for 13 years. The trial has been carried out on a red lateritic soil known as Tippera. The soil is a clay loam with a massive structure, usually well drained and hard setting when it dries (Lucas et al. 1985). The mean annual rainfall is 1224 mm with 91% falling between November and March (Anon. 1987). A variety of legume and poaceous crops have been grown each wet season. In most years the direct drill crops have sustained more insect damage than the conventionally cultivated crops. Most of the insect damage has occurred in the three weeks post germination. The damage was attributed to a variety of pest species, wireworms, false wireworms, crickets and acridid grasshoppers. As a result it was decided to sample the ground dwelling invertebrate fauna in 121 the till and no till plots as well as record the amount of crop damage, one week post germination. Materials and Methods The conventional cultivation/direct drill trial consisted of six plots with three replications. Each replication had a conventionally cultivated plot and a direct drill plot. The plots were 69 m x 54 m. Herbicide was used to kill grasses and dicotyledonous weeds prior to sowing in direct drill plots. Eight pitfall traps were placed in each plot once a month from September until December during the years 1995 to 1997. The traps consisted of polythene take-away food containers, 100 mm in diameter and 110 mm deep. Holes were dug with a mattock and a motorised post hole-digger and the traps were placed in the holes with the lip of the container flush with the soil surface. The traps were placed in a grid pattern, the position of the grid being at random. The grid consisted of two rows with four traps in each row. There was one metre between the rows and traps were spaced at 2 m intervals within the rows. Each trap was filled to one third of its volume with 70 percent ethylene glycol. The traps were placed in the field between 14.00 hrs and 16.00 hrs in the afternoon and collected the next day. Invertebrates were removed from the ethylene glycol solution and placed in glass vials with 80 percent ethanol, before being returned to the laboratory. The catch of insects from each month was sorted to family and in the case of ants to genus and species. Because we were unable to match immature stages with adults, only the adult stage was recorded. A survey of insect damage to the wet season crops in till/no till plots was made one week, post emergence. Surveys were carried out for the 1995/96 and 1996/97 seasons (table 4). Ten rows were selected systematically at 5 m intervals across each plot. A 2 m sample was taken from each row using random numbers, making a total of ten samples per plot. Counts were made of the number of plants emerged and the number damaged by insects. Results Eight species of phytophagous insects were recovered from pitfall traps (Table 1). In general phytophagus insects were more commonly trapped in the no-till plots. The exception being tenebrionids, which were more common from traps in the conventionally tilled plots. The yellowwinged locust, Gastrimargus musicus (Fabricius), was trapped more than four times as often in no-till plots compared with till plots. Gryllids were recovered from no-till plots more than twice as often as from till plots. 122 Table 1. Phytophagous insects recovered from pitfall traps PERCENT OF TRAPS CONTAINING ORDER Coleoptera Orthoptera FAMILY GENUS AND SPECIES FAMILY Till No-Till Elateridae 3 spp. 1.8 1.8 Tenebrionidae 2 spp. 4.2 2.1 Acrididae Gastrimargus musicus (Fabricius) 1 sp. <1 4.2 Gryllidae 2 spp 5.2 11.8 Six species of predacious Coleoptera were recovered from the traps (Table 2). Carabinae were more commonly trapped in no-till plots, while Cicindelinae were trapped twice as often in till as in no-till plots. Cicindelines were more common after rain. One species of Staphylininae was recovered from no-till plots. Table 2. Predacious Coleoptera recovered from pitfall traps Percent of traps containing subfamily Family Subfamily Genus and species Till No-till 4.6 8.8 Chlaenius flaviguttatus Macleay Carabidae Carabinae Clivina spp. Gnathagus pulcher Dejean Harpalina sp. Staphylinidae Cicindelinae 2 15.5 7.9 Staphylininae 1 0 1.8 Twenty-two species of ants were recorded from pitfall traps (Table 3). I. Sanguineus and S. germinata were more commonly recorded from till than no-till traps. R. aurata was more often recorded from no-till than till traps, while R. reticulata and O. nr turneri were equally common from till as no-till traps. The no-till traps yielded a greater diversity of species than the till traps. 123 Table 3. Ant species most commonly recovered from pitfall traps Subfamily Dolichoderinae Myrmicinae Ponerinae Genus and species Number of traps containing species Till No-Till 19.2 7.9 6.7 < 0.1 14.9 27.1 10.3 11.6 7.0 6.1 6.1 12.5 Iridomyrmex sanguineus Forel Solenopsis germinata (Fabricius) Rhytidoponera aurata (Roger) R. reticulata (Forel) Odontomachus nr turneri Forel Other species from various subfamilies The surveys of insect damage showed 9% fewer establishments for mung bean in the no-till plots by comparison with till plots during the 95/96 season, while in the 96/97 season establishment of cavalcade in no-till plots was 33.5 % less than that of till plots. Insect damage was over five times more common on mung bean in no-till plots during 95/96. Damage to cavalcade in the 96/97 season was 30 % higher in no-till plots. Table 4. Mean plants established and percent plants damaged in till and no-till plots Crop 124 Mean plants established per metre Percent plants damaged Till No-Till Till No-Till Date sampled Mung bean 27/12/95 12.0 10.9 3.75 20.09 Centrosema pascorum “cavalcade” 21/1/97 16.07 5.39 12.60 16.40 Discussion Phytophagus insects were trapped more commonly in direct drill than conventionally cultivated plots. The exceptions were the tenebrionids, false wire worms, which were more common in cultivated plots. In view of the prevalence of false wire worm outbreaks in the Douglas Daly and Katherine regions the effect of cultivation on false wire worm populations deserves further investigation. The higher incidence of phytophagus insects in no-till plots compared with till plots, could be explained by the ants, I. sanguineus and S. germinata, being more commonly trapped in till plots. Both species of ant are voracious arthropod predators, which prefer the high temperatures and high light intensity offered by bare cultivated ground (Hölldobler and Wilson 1990; Andersen and Patel 1994). The higher incidence of cicindelines in till and carabines in no-till plots does not have a ready explanation. The higher damage to plants in no till plots confirms previous observations made in the Douglas Daly and Katherine regions. The answer may be that crop residues do not persist in Tippera soils and consequently there is no accumulation of organic matter. High soil temperatures in the region ensure that organic matter is rapidly oxidised (K. Thiagalingam, pers. comm.). Since the litter layer does not build up on the soil surface there is no habitat for decomposers and predatory soil fauna as often happens in temperate soils after the practice of conservation tillage. Acknowledgements Thanks are due to Dr A. Andersen CSIRO, Tropical Ecosystems Research Centre Darwin, for the identification of ants and the staff of the Douglas Daly Research Farm for their help and advice. 125 References Andersen, A. N. and Patel, A. D. (1994). Meat ants as dominant members of Australian ant communities: an experimental test of their influence on the foraging success and forager abundance of other species. Oecologia 98: 15-45. Anon. (1987) Douglas Daly Research Farm Visit. Joint annual DPP/CSIRO/Wada/CCNT. NT Department of Primary Production, internal report. meeting of Cogle, A. L., Bateman, R. J., Heiner, D. H., (1991). Conservation cropping systems for the semi arid tropics of North Queensland, Australia. Aust. J. Exper. Agric., 31: 515 –523. Hölldobler, B. and Wilson, E. O. (1990). The Ants, Springer-Verlag, Harvard, USA. 732 pp. Lal, R., (1986). No-tillage and surface tillage systems to alleviate soil-related constraints in the tropics. In: No-tillage and surface tillage agriculture: the tillage revolution. (Eds Sprague, M. A. and Triplett, G. B.) Pp 261 – 318. Wiley, New York. Lucas, S. J. (1985). Soils of the cleared areas, A.D.M.A. Douglas Daly project farms. Technical memorandum, Land Conservation Unit, Conversation Commission of the NT, Darwin. Robertson, L. N. (1993) Influence of tillage intensity (including no-till) on density of soil-dwelling pests and predatory animals in Queensland crops. In: Pest control and sustainable agriculture, Eds. Corey, S.A., Dall, D. J. and Milne W. M., CSRIO Division of Entomology, Canberra. Pp. 349-352. Robertson, L. N., Kettle, B. A. and Simpson, G. B. (1994) The influence of tillage pratices on soil macrofauna in a semi-arid agroecosystem in northeastern Australia. Agric, Ecosystems and the Environ, 48: 149-156. Stinner, B. R. and House, G. J., (1990) Arthropods and other invertebrates in conservation – tillage agriculture. Ann. Rev. Ent, 35: 299 – 318. 126 DETECTION AND ERADICATION OF THE EXOTIC FRUIT FLY BACTROCERA PHILIPPINENSIS DREW AND HANCOCK (DIPTERA: TEPHRITIDAE) IN THE NORTHERN TERRITORY ESC Smith Entomology Branch Department of Primary Industry and Fisheries GPO Box 990 Darwin NT 0801 Abstract A successful eradication program for the exotic fruit fly Bactrocera philippinensis Drew and Hancock (Diptera: Tephritidae) was mounted between November 1997 and November 1998. At the peak of eradication, about 80 people were employed in protein baiting and caneite block placement. Trap monitoring continued for the full 12 months with a total of 82 B. philippinensis and almost 397,000 other endemic fruit flies collected. The last exotic fly was caught on 26 December 1997, 36 days after the initial detection. Fruit collections were made from all areas within the eradication zones. Over 6,500 samples of fruit were collected from the field over the nine month period to the end of August 1998 while others were selected from fruits confiscated at airport or road checkpoints. No B. philippinensis were reared from any of the fruit collections. Introduction A single specimen of the exotic fruit fly Bactrocera philippinensis was trapped in a permanent Methyl eugenol (ME) trap at Berrimah in the Darwin urban area on 19 November 1997. An additional two flies were caught the following day necessitating the declaration of an outbreak and advising all States and the Commonwealth of the situation on 21 November. On 12 December 1998, a single fly was captured in a ME trap placed on a rural property at Humpty Doo some 30 km east of the original detection, and which had sourced mango fruit from Darwin suburban areas. A successful eradication program was then developed and operated until June 1998. Monitoring the entire quarantine zone by stratified fruit sampling, subsequent rearing of infested larvae and collection of fruit fly adults in lure traps continued until 30 November 1998 when the outbreak was declared eradicated and all quarantine checkpoints removed. Methods The Exotic Fruit Fly Program contained various legal, technical and regulatory components to achieve the aim of eradicating the fly from the NT. These included: • • • • • • • • declaration of a gazetted Quarantine Area delineation of the area infested by B. philippinensis by trapping and fruit collection designation of “eradication zones” erection of checkpoints to prevent infested fruit leaving the Quarantine Area publicity to inform and gain the cooperation of all persons affected by the Program training of primary producers on regulatory requirements imposed on products marketed outside the Quarantine Area regular reporting to Commonwealth and State authorities to ensure cost sharing verification of species eradication by trapping and fruit collection. Methyl eugenol traps were used to detect and monitor adult flies both exotic and endemic species. Almost 400 traps were placed within the Quarantine area and the majority serviced and cleared twice weekly until the end of March 1998. Traps were serviced at weekly intervals thereafter. All traps were global positioned and spatial analysis revealed that the number of traps could be 127 reduced to 290 without compromising the complete coverage of lure traps (Hempel and Roseverne 1998). This was achieved after May 1998. Two eradication zones were designated: one encompassing more than 9000 properties in the Darwin suburban areas of Berrimah, Karama, Malak, Anula, Wulagi, Wanguri, Leanyer, Moil, Jingili, Alawa, Nakara and Wagaman; and the other covering more than 700 rural properties over an area of about 18.5 sq. km in Humpty Doo. The physical eradication strategy had three major elements: • • • removal of as much “windfall” fruit as possible from within the zone to reduce breeding of the fly protein baiting on all private properties and public areas within the majority of the eradication zone at a rate of 2 L/ha placement of Methyl eugenol/maldison impregnated caneite blocks on all areas within the zone at a rate of more than 600 per sq. km. (i.e. >6/ha). Following the arrival of blocks from Queensland, blocking commenced on 04 December 1997 with the aim of placing one caneite block per suburban houseblock or about 8/ha. The blocks were retrieved and replaced with new colour coded ones at 6-7 weekly intervals. Schools and parklands within the area were treated during the school holiday period only. Fruit collections were made from all areas within the eradication zones. These samples were held in a quarantine secure laboratory to enable eggs or larva within the fruit to develop through to adult fruit flies (Luders 1998). Results Following the initial detection, supplementary ME traps were placed in adjacent areas and, within a week, a total of 33 flies had been caught at Berrimah and the adjacent suburbs of Karama and Malak. A total of 82 B. philippinensis flies were trapped in 20 trap locations (Table 1). The majority of these were caught within two weeks in the suburb of Karama. No further exotic flies were detected in Humpty Doo. The last exotic fly was caught on 26 December 1997, 36 days after the initial detection. To the end of November 1998, more than197,000 endemic flies had been collected and examined from the ME traps which monitored adult fly populations. Of these, more than15,000 were B. opiliae (Drew and Hardy), the only endemic species in the Bactrocera “dorsalis” complex, and which needs very detailed examination to separate this species from B. philippinensis. 128 Table 1. Total trap catches of B. philippinensis in the Darwin eradication program 1997 14 S1 20 Nov 21Nov 22 Nov 23 Nov 24 Nov 25 Nov 26 Nov 27 Nov 29 Nov 1 Dec 3 Dec 5 Dec 12 Dec 19 Dec 26 Dec 1 2 2 TOTAL 3 S1 4 5 6 2 1 1 S2 0 S2 1 S2 2 S2 3 1 1 8 1 2 7 7 2 2 1 2 1 15 35 1 1 1 1 S3 8 S1 7 S5 3 S2 8 1 5 2 1 1 1 2 S2 7 S2 4 S2 9 1 1 1 1 1 S5 S12 1 S7 2 S1 2 1 1 1 1 1 3 19 2 2 25 5 2 2 8 1 2 1 1 1 1 1 1 1 Daily Total 1 4 5 6 14 4 1 12 19 7 4 2 1 1 1 Running Total 1 5 10 16 30 34 35 47 66 73 77 79 80 81 82 1 82 Trap Locations:14, S1, = CSIRO S14, S20, S21, S22, S23, 15, S38, S17, S28, S27, S24, S29, = KARAMA 35 = MALAK S53 = ANULA S72 = MOIL S5 = FUNERAL PARLOUR S121 = HUMPTY DOO S12 = BERRIMAH 129 Table 2. Ratified timetable for B. philippinensis eradication program in Darwin February 1998 Increase fruit collection within and outside eradication areas 31 March 1998 Cease protein spot baiting in both eradication areas Mid-April 1998 Cease caneite blocking placement in both eradication zones Commence deblocking in both areas 31 May 1998 Complete caneite deblocking in both areas Maintain intensive trap monitoring and fruit collection Commence introduction of B. philippinensis Area Freedom protocols throughout the Quarantine zone to facilitate interstate trade 31 August 1998 Cease fruit collection 30 September 1998 Complete fly rearing and close fruit handling laboratory 30 November 1998 If no further detections are made, declare full Area Freedom for interstate trade Revoke declaration of Quarantine Area Cease operations at road and airport checkpoints Continue intensive trap monitoring 31 December 1998 Reduce trap monitoring to national grid level 31 May 1999 Declare eradication of B. philippinensis. AQIS will advise other countries of Australia’s freedom from this species. Protein bait applications commenced on 22 November 1997 and most areas were treated at weekly intervals until 31 March 1998. The protein baits (protein autolysate and maldison as the toxicant) were applied in coarse “squirts” of about 40 ml each to 4 or 5 fruiting trees in each suburban houseblock and to many street trees. More than 50,000 caneite blocks were placed during three rounds of blocking. The last round was completed by late April and all areas deblocked by 31 May 1998. Over 6,500 samples of fruit were collected from the field during the nine month period to the end of August 1998 while >300 others were selected from fruits confiscated at the airport or road checkpoints (Luders 1998). No B. philippinensis were reared from any of the fruit collections. 130 A proposed timetable for the eradication program was forwarded to all States for ratification (Table 2) and subsequently accepted. Conclusion The B. philippinensis incursion was detected at an early stage and eradicated from the NT within 12 months. Despite the lack of detection of exotic adult flies in the ME trap array beyond December 1997, fruit monitoring continued until the end of August 1998 and the quarantine checkpoints and trap monitoring continued at a high level until 30 November 1998 to satisfy interstate and overseas trading partners that eradication had occurred. Since no further detections were made, Area Freedom for interstate trade was then declared, the Quarantine Area revoked and all checkpoints closed. Provided no further detections are made, international area freedoms for B. philippinensis should be declared for Australia at the end of May 1999. Acknowledgements I gratefully acknowledge and sincerely thank the numerous personnel appointed to the eradication program who physically conducted the eradication, monitoring, quarantine functions, roadblocks and administration, and without whom the success of the program would have been in jeopardy. References Hempel, CJ and Roseverne, V (1998). Exotic Fruit Fly Eradication Program. Trap monitoring and grid assessment report. Internal Report NT DPIF. 13 pp. + 7 pp. appendices. August 1998. Luders, LG (1998). Exotic Fruit Fly Eradication Program. Fruit handling laboratory procedures report. Internal Report NT DPIF. 18 pp. + 24 pp. appendices. September 1998. 131 DETERMINATION OF HEAT TOLERANCE IN IMMATURE STAGES OF BACTROCERA AQUILONIS (MAY) (DIPTERA: TEPHRITIDAE) FRUIT FLY Heather Wallace Entomology Branch Department of Primary Industry and Fisheries GPO Box 990 Darwin NT 0801 Abstract To enable export of fruit and vegetables from tropical Australia to New Zealand (NZ) strict quarantine regulations are demanded. In order to expand our market to NZ, the NZ Ministry of Agriculture and Fisheries (MAF) has supplied guidelines for disinfestation requirements. Hot water dipping of produce enables reduction in the use of toxic chemicals during processing and packing of produce. Trials on naked fruit fly larvae as well as infested fruit and vegetables are necessary to establish the efficacy of this method of disinfestation. It is a requirement that all stages of fruit fly infestation are controlled before produce will be accepted for export to NZ. The aim of the project was to determine the most tolerant stage of the most tolerant species of fruit fly to immersion in hot water. The Northern Territory fruit fly, Bactrocera aquilonis (May) (Diptera: Tephritidae) is one of the common pest species of quarantine importance in the NT. It was found that 1st instar larvae of B.aquilonis were the most tolerant to immersion in hot water at 46°C and 48°C. Mature eggs followed by 1st instar larvae were most tolerant at 44°C. Young eggs were least tolerant for all temperatures. Trials on other pest species endemic in Western Australia and Queensland were carried out in those states as components in this project. Key words: Bactrocera aquilonis, life stage tolerance to hot water dipping Introduction Post harvest disinfestation of fruit and vegetable produce currently involves the use of chemical agents. In order to reduce the use of toxic chemicals, methods such as hot water dipping are being trialed with good results. This method has been used on a small number of organic mango crops in the Northern Territory. Australia has many pest species of fruit fly that infest economic crops causing products to be unmarketable if protocols for control are not in place. When export of Australian produce to other countries is desired it is essential that quarantine requirements for disinfestation are adequate. New Zealand is a close neighbour of Australia. It is currently free of pest species of fruit fly. This project was set up to comply with the Ministry of Agriculture and Fisheries (MAF) regulations for quarantine of produce being exported to New Zealand from Australia. In order for Australian produce to be accepted we were asked to carry out trials on all major pest species of fruit fly in tropical Australia. The aim of the project was to determine the most tolerant stage of the most tolerant species of fruit fly to immersion in hot water. This should then allow a conclusion that all other stages of all other species less tolerant would be killed under these conditions. The MAF set specific guidelines for disinfestation methods that will be acceptable and these have been incorporated in this project. Concurrent trials for this project were carried out in WA, Queensland., and the NT to gather heat tolerance data on the most prevalent pest species in each area. Work on the commonly occurring fruit fly B. aquilonis was assigned to the Northern Territory. 133 Methods Standardised equipment and methodology were used in all laboratories to establish a high degree of uniformity. Pre-trial testing was carried out to determine the hatch rate and age (h) of hatch for eggs and to establish the number of hours taken to reach 60% development. Six life stages of B. aquilonis were subjected to hot water immersion at three temperatures for varying times. The life stages were: eggs< 2h old, eggs at 60% development, and 1st, 2nd, 3rd and non-feeding 3rd larval stages. Temperatures of 44, 46 and 48°C were used. Each life stage was dipped at each temperature for varying time intervals in order to achieve eight time points per temperature in the range between low and high mortality. The number of insects per dipping tube was 100 for eggs, 100 for 1st instar larvae and 50 for other instars. Control tubes contained insect numbers to match. Five replicates of each stage of development for each temperature were conducted. Two sets of control insects were used for each replicate. Surviving treated and control insects were reared through to pupation and eclosion on carrot media in an incubator set at 26°C. Percentage of mortality was recorded at three phases post treatment: Acute mortality - failure to hatch for eggs at 72h - 24h post treatment for larvae Chronic mortality - failure to pupate Pupal mortality - failure of normal eclosion. Results Young eggs (<2h) were the least tolerant to hot water immersion across all temperature ranges. First instar larvae showed a high degree of tolerance for the three temperatures. The results for each developmental stage at 44°C showed mature eggs as being the most tolerant to immersion with 1st instars being the second most tolerant. At 46°C, 1st instars were the most tolerant followed by mature eggs. Table 1. Tolerance to immersion at three temperatures Most tolerant to least tolerant. 44 degrees C 46 degrees C 48 degrees C Mature eggs 1st instars 1st instars 1st instars mature eggs N/F 3rd instars 2nd instars 2nd instars 3rd instars 3rd instars N/F 3rd instars mature eggs N/F 3rd instars 3rd instars 2nd instars Young eggs young eggs young eggs Discussion: To be useful for disinfestation of fruit and vegetables the temperature of the hot water and the time period used for immersion must not degrade the quality of the produce. 134 At 44°C an immersion time of 40 minutes resulted in 100% pupal mortality for the most tolerant stage (mature eggs). Immersion of produce in hot water for that time is likely to affect the quality of the produce as well as adding to the processing time. At 48°C the immersion time required to achieve 100% pupal mortality for the most tolerant stage (1st instar) was 5 minutes. In this case the time is short however the high temperature of the water may cause damage to the produce. An immersion time of 12 minutes at 46°C achieved 100% mortality at the acute, chronic and pupal phases for the most tolerant stage (1st instar). At this temperature and time, damage to produce should be minimal whilst achieving adequate disinfestation. Final analysis of results and comparison with other fruit fly species are not yet available. A final report will be compiled by Mr R.J.Corcoran, Entomologist QDPI and presented to HRDC and N.Z. MAF by the end of December 1998. Acknowledgement Mr. Stuart Smith, Principal Entomologist, DPIF, NT, for his support and supervision throughout the project. Mr. Robert Corcoran, Entomologist, Queensland DPI, for coordinating the project and providing feedback and support. Miss Leeane Heslin, Queensland DPI, for befriending me during my stay in Indooropilly. Miss Pauline Peterson, Queensland DPI, for her advice and direction in the initial stages of setting up the project. The Project HG619, was funded by the Horticultural Research and Development Corporation (HRDC) with support from the Australian Horticultural Exporters Association (AHEA) and the Queensland Fruit and Vegetable Growers. References Drew R.A.I., Hooper G.H.S., Bateman M.A.(1978) Economic Fruit Flies of the South Pacific Region. Anon. (1994). MAF Regulatory Authority Standard 155.02.03. Specification for the determination of fruit fly disinfestation treatment efficacy. MAF Regulatory Authority, Wellington, NZ. Corcoran, R.J. (1993). Heat-mortality relationships for eggs of Bactrocera tryoni (Froggatt) (Diptera: Tephritidae) at varying ages. Journal of the Australian Entomological Society 32: 307310. Corcoran R.J. (1996). Determination of Heat Tolerance in Immature Stages of Fruit Fly Pest Species. Laboratory guidelines for Project HG619. 135 SIGASTUS WEEVIL - AN EMERGING PEST OF MACADAMIAS IN NORTH QUEENSLAND H.A.C. Fay, S.G. De Faveri, R.I. Storey and J. Watson Queensland Horticulture Institute Queensland Department of Primary Industries PO Box 1054 Mareeba Queensland 4880 Abstract A large weevil was found infesting macadamia nuts on the Atherton Tableland during the 1994/95 season. It was unrepresented in various Australian insect collections but thought to belong to the genus Sigastus. This paper reports some preliminary studies on its biology, pest status and control. From 4-6 weeks after first nut-set adult females commence laying single eggs through the husk, after first scarifying an oviposition site. The nut stalk is then cleaved leading to rapid abscission. Nuts were generally attacked up until hard shell formation. Weevil larvae consumed whole kernels, with % survival higher and larval duration shorter in larger nuts. Infestation rates increased with increasing nut diameter, reaching 72.8% of fallen nuts by midOctober. A crop loss of 30% could be attributed to weevils in an unsprayed orchard. However, adult weevils are very susceptible to both carbaryl and methidathion sprays. In addition, exposure of infested nuts to full sunlight over several weeks kills 100% of larvae. Crops should be surveyed for weevil damage from the 5-10 mm diameter stage until mid-December. Methidathion used as an initial spray for fruitspotting bugs should provide control. Organic growers are advised to sweep infested nuts into mown interrows where solarisation will kill larvae. Introduction The major pests of macadamia in north Queensland include macadamia nutborer, Cryptophlebia ombrodelta (Lower), fruitspotting bugs, Amblypelta spp., and macadamia flower caterpillar, Cryptoblabes hemigypsa Turner (Ironside 1981). Other pests, such as macadamia lace bug, Ulonemia concava Drake, can cause severe damage at times but activity tends to be localised. Weevil pests had been a rare feature of the macadamia industry worldwide, until the appearance of the nut-infesting scolytid, Hypothenemus obscurus (Fabricius), in Hawaii in 1988 (Beardsley 1990). During the 1994/95 season a large weevil was found damaging developing nuts in several commercial macadamia plantations on the Atherton Tableland. The insect was unrepresented in the QDPI Mareeba insect collection, as well as those of the ANIC and the Queensland Museum. The apparently undescribed species is thought to belong to the genus Sigastus Pascoe (Curculionidae: Molytinae: Haplonychini)(Zimmerman pers. com.), and while considered to be native, the possibility that it is a New Guinean species cannot be discounted. The usual hosts of this group of weevils are species of Eugenia and Ficus. This paper reports some preliminary studies on the Sigastus weevil, including its biology, pest status and control. Methods Study sites Studies were conducted on a number of commercial macadamia plantations on the Atherton Tableland, in the Atherton and Kairi areas. Most work was undertaken during the 1997/98 season on the Kairi property of J. Rees, which contained a block of about 1000 mature trees of the varieties 344 and 741. These trees received no pesticide treatments and were not irrigated. Behaviour and biology studies Weevil behaviour was largely observed in the crop during surveys for fruitspotting bug damage. Adult feeding and oviposition was noted. In addition, weevil damaged nuts were collected from the ground to determine any relationship between nut size and weevil emergence rate. On two occasions about six weeks apart 50 nuts were placed individually in 250 ml polystyrene containers, sealed with gauze tops, and held at ambient temperature until adult weevils emerged. 137 Crop damage assessment At fortnightly intervals from nut-set to maturity, nuts beneath 30 trees were sampled for weevil damage. Up to 15 freshly fallen nuts were randomly selected from the ground beneath a tree and the proportion with weevil oviposition scars recorded. Nut diameter was measured with microcallipers. This data provided an estimate of fallen nuts with weevil damage, but not of total crop loss per tree. However, as counts of nuts on panicles were taken before weevil activity commenced, and at the time damage ceased, some crop loss estimates were possible. Virtually all nuts with oviposition scars fell from panicles. Insecticide evaluation Macadamia branches were sprayed in situ with carbaryl, methidathion or β-cyfluthrin and allowed 1-2 hours to dry. The application rates are given in Table 2. All insecticides and the water control were applied with a wetting agent (Agral @ 25 ml/100L). Adult weevils then were enclosed in 1 m long gauze sleeves over the treated branches for 24 hours. There were 5 weevils/sleeve and 6 replicates of each treatment. After 24 hours the weevils were assessed as alive or dead, and later at 72 hours after holding in 250 ml polystyrene containers. Results Behaviour and biology Female weevils began laying eggs in nuts about 4-6 weeks after first nut-set. The female scarifies an area about 3-4 mm in diameter in the husk and lays a single egg into it. The egg is either lodged within the husk or intrudes into the surface of the kernel. After oviposition the nut stalk is chewed about half through to induce drop. Generally, a single egg was laid per nut, although up to three eggs (or larvae) were observed in a nut representing different oviposition dates. Although a large proportion of nuts within a panicle could be affected it was rare to find that all were. Nut fall was rapid and generally occurred from a few day after oviposition. Mating was observed on nut panicles. After nut shells hardened (about mid-December) and were no longer suitable for oviposition, adult weevils commenced to feed on the green surface of the husk, in some cases completely removing the epidermis. Adults will also feed on young leaves. Larvae consumed entire kernels before pupating and chewing exit holes as adults. Table 1 provides details of adult emergence based on different nut size ranges. Survival was higher and average development time was shorter in larger nuts. Larger weevils emerged from larger nuts, as reflected by emergence hole diameters. The minimum development time in nuts was about 6 weeks, but nuts were not retrieved from the time of oviposition to confirm this. Table 1. Some relationships between macadamia nut size and utilisation by Sigastus Collection date 17 Oct. 1997 26 Nov. 1997 Average nut diameter (mm) 16.9 28.1 Nut size range (mm) 14.3-21.8 22.5-32.3 % adult emergence 29.0 40.6 Emergence hole diameter (mm) 5.5 6.7 Maximum time to emergence (days) 80 49 Damage assessment The first weevil-damaged fallen nuts were collected on 18 September 1997 (Figure 1). Infestation rates increased with increasing nut diameter, reaching 72.8% of fallen nuts by 17 October. The infestation rate remained at about this level for about four weeks before declining in December. Nuts reached their full size by early December, with shells hardening soon after. The number of fallen nuts increased rapidly through September and early October, then levelled-off before declining in early December. 138 80 120 100 60 80 50 40 60 30 40 20 20 F allen n uts (% of ma x. s am ple ) % infeste d n uts / av era ge nu t dia m eter 70 10 0 0 3/09/97 18/09/97 2/10/97 17/10/97 30/10/97 14/11/97 28/11/97 12/12/97 % infested nuts (ground) Average nut diameter (mm) Fallen nuts (% of max. sample) Figure 1. Rate of infestation of macadamia nuts by Sigastus and nut fall induced in relation to crop development Insecticide evaluation Both carbaryl and methidathion caused 100% mortality of adult weevils within 24 hours of application (Table 2). β-cyfluthrin was slower acting, with some repellency indicated, but a high level of mortality was recorded after 72 hours. Table 2. The effect of 3 different insecticides on mortality in adult Sigastus Treatment (+ wetting agent) Control Carbaryl Methidathion β-cyfluthrin Product/100L 125 g 125 ml 50 ml Post-application % mortality 24 h 72 h 0.0 10.0 100.0 100.0 100.0 100.0 46.7 86.7 Discussion Sigastus weevil can cause significant crop loss in macadamias where insecticidal control is not practised. Although damage to nuts in individual trees varied greatly, an overall crop loss of around 30% was estimated at the monitoring site. This weevil could be a significant pest for organic growers, or those using minimal insecticidal treatments. However, trials in which infested nuts were exposed to full sunlight over several weeks indicated that solarisation caused 100% mortality of larvae. Heat treatment has previously been suggested for dealing with larvae of acorn weevils in India (Kaushal et al. 1993). For growers engaging in normal management practices, existing chemicals can control adult weevils. Methidathion, employed as an initial spray for fruitspotting bug, coinciding with the very first sign of nut drop, should provide effective control of Sigastus weevil. Follow-up sprays for nut borers would ensure further control. Nuts falling between mid-September and mid-December should be swept into mown interrows where solarisation will kill larvae. However, it appears that Sigastus may reoccur in a crop each season, irrespective of the control measures adopted previously. Weevil damage to nuts is 139 obvious and crops should be surveyed thoroughly when nuts reach between 5 and 10 mm in diameter. Because weevils only appear to oviposit in nuts prior to shell hardening it seems highly unlikely that adults would emerge from mature nuts. This suggests that the chance of infested nuts being moved from the Atherton Tableland to southern shellers is negligible. However, the adults are robust and could inadvertently be moved amongst bulk nuts. The species will undoubtedly remain a nuisance to northern growers, and vigilance will be required to minimise its impact. Acknowledgment John Rees is particularly thanked for access to his macadamia crop. The Queensland Horticulture Institute has authorised publication of this paper. References Beardsley, J.W. (1990). Hypothenemus obscurus (Fabricius) (Coleoptera: Scolytidae), a new pest of macadamia nut in Hawaii. Proc. Hawaii Entomol. Soc. No. 30: 147-150. Ironside, D.A. (1981). Insect pests of macadamia in Queensland. QDPI Misc. Publ. 81007. 28pp. Kaushal, B.R.,Pant, M.C., Kalia, S, Joshi, R. and Bora, R. (1993). Aspects of the biology and control of three species of acorn weevils infesting oak acorns in Kumaun Himalaya J. Appl. Ent. 115: 388-397. 140 NEW AND POTENTIAL ARTHROPOD PESTS RECORDED IN THE NORTHERN TERRITORY FROM 1991 - 1997 E.S.C. Smith and H.H. Brown Entomology Branch Department of Primary Industry and Fisheries GPO Box 990 Darwin NT 0801 Abstract This paper records the detections of new or potential arthropod pest species in the Northern Territory between 1991 to 1997, which period corresponds to the staging of the 5th and 6th Tropical Entomology Workshops. In addition, records for the significant expansion of known pests and brief notes on each species mentioned are provided. The scope and frequency of these new records signifies both the growth of the NT horticultural and agricultural industries with associated pest development derived from interstate and local bush sources and the very real risks involved with exotic arthropod incursions into the sparsely populated northern coastline of Australia. Pest Records 1991 Severe defoliation of rambutans by the castoroil looper caterpillar Achaea janata (Linnaeus) (Lepidoptera: Noctuidae); New fruit fly pest, Bactrocera cucumis (French) (Diptera: Tephritidae) found damaging cucurbits in the Top End in a short-lived outbreak; The newly described (from cashew) sucking bug Helopeltis pernicialis Stonedahl, Malipatil and Houston (Hemiptera: Miridae) detected on mango trees in both Darwin and Katherine; Swarming Graptostethus servus (Fabricius) (Hemiptera: Lygaeidae) on bananas caused physical blemishes; Cyrtopeltis sp. (Hemiptera: Miridae) causing considerable injury to sesame at Katherine; Nezara viridula (Linnaeus) (Hemiptera: Pentatomidae) in high numbers and damaging carambola fruit at Coastal Plains Horticultural Research Station; First Australian record of Batrachedra sp. (Lepidoptera: Batrachedridae) collected on coconut flowers in Darwin. 1992 First record of longicorn beetle, Acalolepta mixtus (Hope) damage (Coleoptera: Cerambycidae) to seedling mangoes; Oviposition by cicadas possibly Illyria hilli (Ashton) (Hemiptera: Cicadidae) caused lesions on seedling mango trees; New record of Diocalandra frumenti (Fabricius) (Coleoptera: Curculionidae) breeding on ornamental Phoenix roebellini palm; First observed damage to guava by H. pernicialis Stonedahl, Malipatil and Houston (Hemiptera: Miridae); 141 Severe two-spotted mite Tetranychus urticae Koch (Acarina: Tetranychidae) damage to papaws; Damage to ripening figs by Carpophilus sp. (Coleoptera: Nitidulidae) (dried fruit beetle) at Ti Tree; First Australian record of the mite Cisaberoptus nr. kenyae Keifer (Acarina: Eriophyidae) from mango leaves in Darwin. 1993 False wireworm, possibly Gonocephalum sp. (Coleoptera: Tenebrionidae) damage to reduced tillage sorghum on Douglas-Daly farms; Major outbreaks of yellow-winged grasshoppers Gastrimargus musicus (Fabricius) (Orthoptera: Acrididae) in the northern pastoral zone; First detection of banana rust thrips Chaetanaphothrips signipennis (Bagnall) (Thysanoptera: Thripidae) in Northern Territory bananas. 1994 Seed crops of pasture legumes damaged by tomato thrips Frankliniella schultzei (Trybom) (Thysanoptera: Thripidae); First record of moth larvae, possibly the native Moroga sp. (Lepidoptera: Xylorictidae) damaging mango trees; First detection in Australia of poinsettia whitefly Bemisia tabaci Type B (Gennadius) (Hemiptera: Aleyrodidae). The pest was severely damaging a rockmelon crop and had been introduced to the NT on ornamental plants from the Eastern States; New record of Brevipalpus mite (Acarina: Tenuipalpidae) blemishing banana fruit; Heavy infestations of Icerya aegyptiaca (Douglas) and I. seychellarum (Westwood) (Hemiptera: Margarodidae) on mangoes, other fruit trees and ornamentals in the Darwin and Darwin rural areas; Severe but localised sucking bug, H. pernicialis Stonedahl, Malipatil and Houston (Hemiptera: Miridae) infestations on bearing mango trees. 1995 Reduced tillage seed establishment problems on Douglas-Daly farms; Damage to pangola pastures by Oulema rufotincta (Clark) (Coleoptera: Chrysomelidae); Green house whitefly Trialeurodes vaporariorum (Westwood) (Hemiptera: Aleyrodidae) detected for first time in NT on imported ornamentals. By May 1995 it appeared to have established in the field; Red banded thrips Selenothrips rubrocinctus (Giard) (Thysanoptera: Thripidae) detected in field planted mangoes at Katherine for the first time; New species of Pseudodendrothrips darci (Girault) (Thysanoptera: Thripidae) severely damaging banyan trees up to 150 years old in Darwin; First detection of coconut aphid Cerataphis palmae (Ghesqhuiere) (Hemiptera: Aphididae) in Darwin; 142 The native wasp Euryglossina melanocephala Exley (Hymenoptera: Colletidae) caused significant damage to stems of commercial Kangaroo paw in Central Australia. 1996 Severe infestations of root nematode Meloidogyne spp. on snake beans; Oviposition damage caused by the cicadas Abricta castanea Distant and Illyria hilli (Ashton) (Hemiptera: Cicadidae) to citrus near Darwin. Later, severe damage also detected at Katherine; New mango problem - eventually determined as mango scab Elsinoë mangiferae; Carpenter bees (Hymenoptera: Xylocopinae) attacked soursop trees; Severe defoliation by Ryparida sp. (Coleoptera: Chrysomelidae) to Jackfruit; New record of Thrips hawaiiensis (Morgan) (Thysanoptera: Thripidae) in mango flowers. 1997 New spider mite recorded on dates at Alice Springs. Subsequently described as Oligonychus calicicola Knihinicki and Flechtmann (Acarina: Tetranychidae); First detection of Oriental red mite Eutetranychus orientalis (Klein) (Acarina: Tetranychidae) in the NT from passion fruit vine, Heliconia leaves and carambola leaves; Attack to aerial roots of commercial sorghum by Euproctis sp. (Lepidoptera: Lymantriidae) in association with Amata sp. (Lepidoptera: Arctiidae); Renewed complaints by householders bitten by psocids at Tennant Creek; Coconut whitefly, Aleurodicus destructor Mackie (Hemiptera: Aleyrodidae) collected from water apple (NAQS detection); Foxtail palms killed by combined attack of Tirathaba rufivena (Walker) (Lepidoptera: Pyralidae), Xylotrupes gideon (Linnaeus) (Coleoptera: Scarabaeidae) and Diocalandra frumenti (Fabricius) (Coleoptera: Curculionidae); Mangoes defoliated by a new swarming beetle pest, Agetinus aequalis Blackburn (Coleoptera: Chrysomelidae); Commercial lemon grass crop damaged by the swarming linear bug, Phaenacantha australiae Kirkaldy (Hemiptera: Colobathristidae); First NT detection of oriental scale, Aonidiella orientalis (Newstead) (Hemiptera: Diaspidae) on Dendrobium orchids; Waterlily aphids, Rhopalosiphum nymphaeae (Linnaeus) (Hemiptera: Aphididae) on Nelumbo nucifera reached damaging levels and required control; Severe infestation by two scales Pinnaspis strachani (Cooley) (Hemiptera: Diaspididae), Parasaisetia nigra (Neiter) (Hemiptera: Coccidae) and a mealybug Maconellicoccus hirsutus (Green) (Hemiptera: Pseudococcidae) on commercial asparagus at Katherine; Red banded thrips (Selenothrips rubrocinctus (Giard) (Thysanoptera: Thripidae)) damaging rambutan fruit and Heliconias; New thrips (Dicromothrips corbetti (Priesner) (Thysanoptera: Thripidae) damaging Vanda orchid flowers; 143 First detection of mango leafhopper, Idioscopus nitidulus (Walker) (Hemiptera: Cicadellidae) in Australia from Darwin suburbs. Very shortly after, this species was detected on Cape York, North Queensland; New gall wasp (Hymenoptera) attacking and killing banyans (Ficus virens) in Darwin; Detection and successful eradication program mounted against the exotic fruit fly Bactrocera philippinensis Drew and Hancock (Diptera: Tephritidae); Diocalandra frumenti (Fabricius) (Coleoptera: Curculionidae) extracted from severely distorted emergent shoots of Royal Palm. Acknowledgment Records come from a range of collectors and the species were identified by various taxonomists. We are grateful to them all. 144 BIOLOGICAL CONTROL OF THE GIANT SENSITIVE PLANT WITH HETEROPSYLLA SPINULOSA (HOM. : PSYLLIDAE) IN PAPUA NEW GUINEA L. S. Kuniata and K. T. Korowi Ramu Sugar Limited P.O. Box 2183 Lae Papua New Guinea Abstract Psyllid, Heteropsylla spinulosa was first introduced in 1991 and now has established in Papua New Guinea and good control of Mimosa invisa has been achieved from 1993 to 1998 seasons. Although the 1997 drought severely affected H. spinulosa populations, these quickly reestablished and provided good control in 1998. Application of nitrogen to Mimosa plants significantly increases the psyllid numbers. Large numbers of the psyllids (>1,000/30 cm tip) are required for traditionally M. invisa problem areas before or at the onset of the wet season. The early establishment of the psyllid in plants thus has the potential of inflicting the greatest damage. Cost of herbicides has been reduced and areas once under M. invisa have been reclaimed with up to 20% increase in calving rates observed at Ramu Sugar Ltd. However, the introduction of the parasite for the biological control of Heteropsylla cubana is a concern to this program. Keywords Mimosa invisa, Heteropsylla spinulosa, biological control, Papua New Guinea. Introduction Mimosa invisa Mart. ex Colla (Mimosacease) commonly known as giant sensitive plant (GSP), has become a serious weed in many parts of South East Asia and the Pacific Islands including Papua New Guinea (PNG) and Australia (Holm et. at. 1977). It is now widespread in coastal and island areas and spreading into the Highlands, and has become a major weed of agriculture, pastures, wastelands and roadsides (Verdcourt 1979). It is difficult to estimate economic losses and the cost of control of GSP for the whole of PNG. However, Kuniata (1994) reported from cattle properties owned by Ramu Sugar Ltd in the Ramu/Markham Valleys in Madang/Morobe Provinces that up to US$130,000 annually was spent on the chemical control and slashing of this weed. On its sugarcane estate, up to threeengine hours down time per day per harvester was experienced as a result of GSP interference with normal sugarcane harvesting (green cane). Persistent herbicides such as 2, 4-D used for GSP control are not only a hazard to people handling them but also can contaminate the environment and cause pesticide residues in animal products. The control of such widespread weed species like GSP requires sustainable efforts and constant supply of resources, which are often limited. Biological control offers sustainable control and also is safe to people and the environment, however, as a pre-requisite to a successful program, the biology and ecology of both the agent and target weed species need to be adequately established. The psyllid (H. spinulosa) is a native of Central America and is probably confined to M. invisa as a host plant (Muddiman et. al. 1992). From about 100 plant species tested by Wilson and Garcia (1992), H. spinulosa developed successfully only on M. invisa indicating its high specificity to this host. Kuniata (1994) observed that application of nitrogen to GSP indirectly increased the psyllid populations and severe damage was inflicted on the plants. In this paper, we report on the establishment and impact of Heteropsylla spinulosa Muddiman, Hodkinson and Hollis (Homoptera : Psyllidae) for the biological control of GSP in PNG. 145 Materials and Methods Most of the field trials and observations on H. spinulosa was made at Ramu Sugar Estate, Gusap, Papua New Guinea between 1991 and 1998. The number of sites used varied each season. As a result of some stations were destroyed before these could be sampled. The psyllid numbers were estimated from the top 30 cm tip of GSP randomly sampled from a clump. Number of eggs/nymphs found on leaf numbers 4 and 6 (unopened leaf tip as leaf number 1) were used to estimate total psyllid numbers on the 30 cm tip. Ground cover by GSP was subjectively scored between 1 (no GSP) and 6 (75 - 100% ground cover). Average clump height was measured from the centre using a metre ruler. The number of clusters was estimated from five randomly sampled 1 m x 1 m quardrats taken from within a clump. Then within each quardrat, all the clusters in a 25 cm x 25 cm area was harvested and number of pods and seeds counted. In an attempt to demonstrate the effect of H. spinulosa damage on GSP, a field trial was established at Gusap ranch in 1994. Infested cuttings of GSP from breeding sites and distributed in the selected clumps. A total of 20 plots each consisting an area between 5-12 m2 were used (depending on clump size) where 10 plots were regularly sprayed with actellic and not the others. Psyllid numbers were estimated one month after introduction. When the seedpods had matured, the seed production was estimated from each using a similar procedure as previously described. Two field cages (1 m x 1 m x 1 m) were set up and in one cage, 50 grams of urea was applied to four week-old GSP. The psyllids were put on plants under these cages and observed for five weeks for the effect of nitrogen. Sixteen M. invisa clumps (9-26 m2) in the field were used to study the effect of nitrogen application on populations of H. spinulosa. Eight of these clumps received 500-600 grams of urea while the others were used as controls (unfertilised). Nymph numbers were estimated from 10 x 25-30 cm tips one month following applications of urea and insect releases. The data were analyzed using paired t-test. Results Field releases of H. spinulosa The psyllid was first introduced into PNG from Tropical Weeds Research Centre, Charters Towers, Queensland in April 1991 and was quarantined at Lowlands Agricultural Research Station, Kerevat, East New Britain Province. A second consignment was received in the same station in November 1991. However, both consignments failed in quarantine as a result of contamination of breeding cages by the fungus Verticillium sp. The third consignment of more than 1,000 insects was received at Laloki Research Station in early December, 1992 and were successfully reared through one generation in quarantine. By the end of December, more than 1,000 F1 generation insects were sent to Ramu Sugar Ltd., Gusap (RSL), and a further 900 insects were received in January, 1993. Field releases of H. spinulosa began in late February, 1993 and continued until June with a total of 250,000 insects released into 36 sites around RSL estate and Gusap ranch (14 km away from RSL). By August 1993 the psyllid was well established in the field and could be found 2-3 kilometres downwind from point of release. The dry season at RSL normally begins in May and ends in September. Most GSP at RSL flower in April-May and start seeding in June-July. This adversely affected the psyllid's activities in the rangelands and attempts were made to maintain the populations on young irrigated plants elsewhere on the RSL estate. In October 1993 the GSP growing along the Ramu River was inspected and it was found that the psyllid was in larger numbers. By December, when GSP started growing vigorously in the rangelands and sugarcane fields, the psyllids immediately moved into these areas. The same thing happened in 1994 and 1995 seasons, indicating that the psyllid is firmly established in this area and can spread naturally. In early 1994, the psyllid was recovered at Brahman High School, some 80 km to the west of RSL. Further releases were made at Yonki dam area in the Eastern Highlands and Leron and Munum Ranches in the Morobe Province. In all these sites, the psyllids established and had an immediate impact on GSP. 146 In 1996 and 1997, the psyllid continued to provide good control of GSP in all the monitoring sites at RSL estate and Gusap ranch (Table 1). However, the psyllid's impact on GSP in 1998 was delayed as a result of poor recolonisation after the prolonged drought in 1997. Although the control was poor, this was still better than the infestations observed pre-1992. In October 1998 most of these sites were again visited and large numbers of psyllids were found on mature and seedling GSP reducing them to a non-significant weed in pastures. In Lae where the climatic conditions are humid with no distinct dry season, the drought of 1997 had no significant impact on psyllid numbers and therefore control of GSP was excellent. In monitoring sites at RSL, the psyllid has had a spectacular impact on GSP since its release in the field (Table 1) reducing it to an insignificant weed in pastures and sugarcane fields. The greatest damage was observed in the seedling stage to 8-10 weeks old GSP where infested plants die prematurely without producing any seeds. This allowed pastures and other shrubs to come through amongst the dying GSP and eventually taking over the area. We found that in order to stop 10-18 weeks old GSP from seeding, the psyllid population had to be greater than 1000/30 cm tip (r =-0.658, p<0.001, df = 39). This population should be established well before the beginning of the wet season. Despite heavy damage in mature GSP (>3 months), some plants will still produce seeds, however, this will be significantly lower (p<0.001) than unattacked plants (pre-1992). The damage is more apparent at the onset of the dry season (moisture stress). Table 1. The effect of H. spinulosa on GSP seed production at Ramu, Papua New Guinea; field releases of the psyllid began in February, 1993 YEAR 1991 1992 1993 1994 1995 1996 1997 1998 No. of samples 40 40 75 44 51 36 43 43 Ground cover score* 5.9 5.8 5.2 3.4 1.8 1.6 1.8 3.8 Avg. clump height (m) 0.5 0.7 0.4 0.4 0.3 0.3 0.2 0.4 No. of clusters/m2 300.2 242.0 126.4 57.2 50.0 106.0 31.0 128.0 55.4 61.4 36.2 9.2 8.3 7.0 10.0 26.0 16,630 14,860 4,580 530 415 439 326 3,546 No. of seeds/cluster Est. no. of seeds/m2 * Score: 1 = 0-1%, 2 = 1-5%, 3 = 5-25%, 4 = 25-50%, 5 = 50-75%, 6 = 75-100% ground cover. In the psyllid exclusion experiment it was difficult to keep the GSP plants insect free, however populations on non-sprayed plots were significantly higher than sprayed ones (Table 2). Ground cover, seed cluster numbers and seed production were highly significantly lower (p<0.001) than sprayed plots further indicating the potential of the psyllid in the control of GSP at RSL. 147 Table 2. Populations of H. spinulosa and seed production in insecticide sprayed and nonsprayed GSP at Gusap Ranch. Number of plots in each treatment was 10. (Ground cover score is similar as for Table 1) Sprayed Non-sprayed t-test Ground cover score 3.1 2.0 9.22 *** No. of clusters/m2 238 47.0 19.19 *** No. of seeds/cluster 56.3 10.5 63.95 *** 13,400 494.1 484.32 *** 309 1136 194.68 *** No. of seeds/m2 No. of psyllids/25 cm tip *** means p<0.001 Distribution of H. spinulosa in PNG The psyllids were mass reared in plots near the laboratory and also in several places on the estate and Gusap Ranch. It was from these plots that insects were collected and distributed to other parts of PNG. GSP tips or portions of 25-30 cm length infested with the psyllids (preferably 4-5 instars) were cut and packed into cardboard boxes (approx. 40 cm x 30 cm x 30 cm) and ice blocks were used to chill the insects during transit. This was sufficient to keep the insects alive for up to 36 hours. Up to 1 kg of urea fertiliser was also packed with the insects so that this could be used to fertilize GSP before insects were released in the field. 148 Figure 1 shows provinces where the psyllids have been introduced since 1993 and except for East and West New Britain provinces, the insects have established and excerting very good control in all the release sites. The results have been quite spectacular in the Madang and Morobe provinces and recently from New Ireland province (R. D. Thorold, personal communication). Figure 1. Releases and Establishments of H. spinulosa in PNG Nitrogen application In the cages, populations of H. spinulosa in urea treated M. invisa plants increased rapidly reaching highest levels in early January, 1992 while those in the control showed a gradual increase over a five week period (Figure 2). By 27 December, the damage in M. invisa was very severe in urea-treated plants. Damage in the control cages was severe as from 7 January 1992. These observations were also repeated in strip trials in the field. Clumps of M. invisa where urea had been applied showed significantly higher numbers of nymphs compared with the controls (t = 4.20, p < 0.01, df = 8). Damage in the urea treated clumps was also severe and M. invisa was reduced to become a less significant weed in pastures two months after treatment. 149 800 Nymphs per 30cm tip (number) 700 600 500 400 300 200 100 0 13-Dec-91 20-Dec-91 27-Dec-91 Control 03-Jan-92 07-Jan-92 Urea Figure 2. Populations of H. spinulosa on M.invisa plants treated with urea. About 500 nymphs were released in each cage and these were allowed to breed Discussion The results from monitoring showed the psyllid had established and was spreading naturally from release sites providing excellent control of GSP. It is quite difficult to quantify accurately the benefits of this program, however, the greatest gains have been realised on cattle properties. On RSL ranches, all the areas that once used to be under GSP (40%) have now been returned to pastures and up to 20% increase has been observed in calving rates (RSL internal report). The herbicide 2, 4-D is no longer used for the control of GSP thus saving on chemical costs. These costs would have escalated annually had it not been for the introduction of the psyllid. And also in sugarcane, no chopper harvester downtime was experienced from 1994 to 1997 seasons in the harvesting of traditionally GSP infested blocks at RSL. The impact of the psyllid has significantly reduced the competitiveness of GSP resulting in this weed attaining a minor status in Papua New Guinea. However, in prolonged dry seasons (e.g. 1997), psyllid numbers were low and it took up to three months before the impact could be realised. Although it is difficult to predict such events, it would be necessary to maintain psyllid populations during the dry season with irrigated GSP or use plants growing along river banks. Application of nitrogen has been shown here to increase psyllid numbers on fertilised GSP which could be treated as part of the routine pasture management. The previous infestations of GSP in PNG had produced a lot of seeds and these will take quite a while to be depleted. However, the introduction of parasitic wasps, Psyllaephagus spp (Encyrtidae) in Indonesia for the control of Heteropsylla cubana Crawford is of major concern to Papua New Guinea since we share the same land mass. Tolerance of Leucaena leucocephala to H. cubana damage should be exploited rather than use parasites, which will only jeopardize the GSP biocontrol program. The results from small cage trials with predatory insects such as lady bird beetles, lacewings, larvae of syrphid flies and spiders were not conclusive. However, 150 the impact of these predatory insects in the field usually come after the psyllids have caused substantial damage to GSP and therefore do not pose a real threat. References Holm, L.G., Plucknett, D.L., Pancho, J.V. and Herberger, J.P. (1977). The world's worst weeds. Distribution and Biology. Univ. Press of Hawaii, Honolulu, 609 pp. Kuniata, L.S. (1994). Importation and establishment of Heteropsylla spinulosa (Homoptera:Psyllidae) for the biological control of Mimosa invisa in Papua New Guinea. International Journal of Pest Management 40 (1) : 64-65. Muddimum, S.B., Hodkinson, I.D. and Hollis, D. (1992). Legume-feeding psyllids of the genus Heteropsylla (Homoptera:Psyllidae). Bulletin of Entomological Research. 82:73-117. Verdcourt, B. (1979). A manual of New Guinea Legumes. Botany Bulletin No. 11. Division of Botany, Lae, Papua New Guinea. 645 p. Wilson, B.W. and Garcia, C.A. (1992). Host specificity and biology of Heteropsylla spinulosa (Hom.:Psyllidae) introduced into Australia and Western Samoa for the biological of Mimosa invisa. Entomophaga 37(2) : 293-299. 151

Log In

INSECT FAUNA SURVEYS ON RAMBUTAN, DURIAN AND MANGOSTEEN IN NORTH QUEENSLAND

INSECT FAUNA SURVEYS ON RAMBUTAN, DURIAN AND MANGOSTEEN IN NORTH QUEENSLAND

Related Papers

RELATED PAPERS