Biological control agents for the guava moth, Coscinoptycha improbana

Sustainable Farming Fund Project L11/145

J. J. Dymock

July 2012

Co-funded by the following Industry Groups and Regional Councils

Project Team Composition

Project Manager: Dr J J Dymock Member 1: (Chair) Don McKenzie, NRC Biosecurity Senior Programme Manager Member 2: Tim Harper, Feijoa grower (Maungataroto) and President of NZ Feijoa Assn Member 3: Dr Lisa Jamieson, Plant and Food Research Member 4: Dr Mike Butcher, NZ Pipfruit Industry Member 5: Dr Chris Hale, NZ Summerfruit Industry Member 6: Bev Worthington, President, NZ Macadamia Society

Growers: Jeff Cleghorn, Macadamia Grower (Kerikeri) Murray Redpath, NZ Tree Crops Assn, Peter Albertson, Feijoa Grower (Kerikeri) Regional Councils: Don McKenzie - Northland Regional Council, Jack Craw – Auckland Council, Kevin Collins -Waikato Regional Council, John Mather- Environment Bay of Plenty

SUMMARY

A search for potential biocontrol agents for the guava moth, Coscinoptycha improbana, was undertaken in urban Sydney, Australia, in March 2012. Guava moth was reared from fruit of lilly pilly, Syzygium paniculatum and S. australe trees from 23 of 28 collection sites sampled. At least five species of wasps that emerged from the fruit samples are potential parasitoids of guava moth. These included two species of Ichneumonidae (one of which is tentatively identified as in the genus Eriborus) and at least three other wasps species from the family Braconidae. Other potential hosts of these parasitoids which emerged from the fruit samples included Sigastus spp. weevils, Bactrocera sp. fruit flies and leafroller moths (Tortricidae). Cytochrome oxidase I gene sequences from guava moth collected from New Zealand and Australia during the project showed that the guava moth from both countries were the same species. Of the guava moth samples analysed, two haplotypes were found in New Zealand; one different from the Sydney specimens, and the other haplotype the same as some of the specimens from Sydney.

INTRODUCTION

Guava moth, Coscinoptycha improbana (Carposinidae), is a native of Australia, where it is a not a pest. It is found in Queensland, New South Wales, Victoria and as far south as Tasmania. In Australia, the larvae have been found boring into the fruits of native and exotic species, Schizomeria ovata (Cunoniaceae), Citrus spp. (Rutaceae), Cassine australis (Celastraceae) and Psidium (guava) sp. (Myrtaceae) (Common 1990).

In the Australian National Insect Collection (ANIC), material of C. improbana is mostly from rain forest habitats from the Atherton region to the Bateman's Bay area, but there is also a specimen from Black Mt., Canberra. One specimen has been reared in August/September from Acmena smithii (Myrtaceae) fruit at Wingham, NSW (Marianne Horak pers. comm.). Juniper and Britton (2010) also report specimens in the ANIC have been reared from: magenta lilly pilly, Syzygium paniculatum, brush cherry, S. australe (Myrtaceae), Acmena ingens (previously Eugenia brachyandra), and the rainforest plant, Elaeodendron australe (Celastraceae). Host fruit of guava moth specimens in the Australian Museum collection also include blue lilly pilly, Syzygium oleosum. In New Zealand lilly pilly are sometimes collectively known as known as monkey apple, and are often confused with acmena.

Guava moth is a pest of soft fruit in Norfolk Island, affecting commercial peach, Prunus persica (Rosaceae), production (Albert Buffett pers. comm.). In New Caledonia, C. improbana has been reared from berries of the endemic shrub Eugenia hurlimannii (Myrtaceae) in the south of the main island, Grande Terre (Mille et al. 2012). In Australia, the endemic plant Eugenia reinwardtiana is the sole member of this genus, although many species in the genera Acmena and Syzygium were once classified under Eugenia. E. reinwardtiana is found in coastal tropical and subtropical Queensland but it is not known whether it is a host of C. improbana.

In New Zealand, guava moth is found in the northern North Island, as far south as Waikato. Its larvae infest a range of soft fruit and nuts year-round, including: citrus, feijoa (Acca sellowiana, Myrtaceae), peaches, plums (Prunus spp. Rosaceae), pears and nashi (Pyrus pyrifolia, Rosaceae), guava, macadamia (Macadamia spp. Proteaceae) and loquat (Eriobotrya japonica, Rosaceae) (Jamieson et al. 2004). The larvae develop within the fruit, so insecticidal control options are limited. Other management techniques for this pest – such as mating disruption and removal of windfall fruit –

have failed to control guava moth. There is continuous re-infestation of fruit from breeding populations of guava moth in a range of environments, including: home gardens, derelict orchards, roadsides and hedges. No natural control agents of guava moth have been recorded in New Zealand. It is likely that guava moth is kept at low population numbers in Australia and New Caledonia by competition with other fruit feeders and/or by predation or parasitism. Biological control may be the most cost effective long-term, sustainable method of control for this pest in New Zealand. In this report, the discovery and identification of potential biocontrol agents of guava moth in Sydney, Australia is documented.

DNA analysis of guava moth populations in New Zealand, Australia and New Caledonia was undertaken to help determine the source and confirm the identity of guava moth in New Zealand as the target for a biocontrol programme.

AIMS

To search for potential biocontrol agents of guava moth in Australia and compare the DNA of populations of guava moth from New Zealand, Australia and New Caledonia.

METHODS

Preliminary pheromone trapping Pheromone trapping of guava moth was undertaken in November 2011 at:

• Bundaberg, QLD

• Alstonville, northern NSW and

• Gosford, NSW. Delta traps baited with DESIRE 2-component guava moth pheromone (Gibb et al. 2006) were placed one trap per tree in five trees in each of these three areas. Only one guava moth was trapped – at Bundaberg, in November 2011. It was decided that the search for biocontrol agents would be re-directed to Sydney where lilly pilly fruit (Syzygium spp.) were recorded as hosts of C. improbana in March 2009 (Juniper and Britton, 2010).

Collection of guava moth in Sydney On 20 and 22 March 2012, fruit of magenta lilly pilly, Syzygium paniculatum (Myrtaceae) and brush cherry, S. australe and S. australe hybrids (collectively referred to as “lilly pilly” in this report) were collected from inner-west and northern Sydney suburbs (see map). The lilly pilly fruit were incubated on a bed of untreated sawdust in 2L plastic “ice-cream” containers and sealed with a cloth mesh lid for 30 days at room temperature in a laboratory at the University of Sydney.

Map of Sydney suburbs where samples of lilly pilly were collected.

Photo: Magenta lilly pilly (Syzygium paniculatum) incubated in 2L plastic containers.

10 samples

7 samples

11 samples

Photo: Samples of lilly pilly fruit incubated in 2L containers in the laboratory at the University of Sydney Over the period 23–27 April 2012, the contents of each container were examined. Any insects present had emerged from the lilly pilly fruit since the experiment had been set up (20 and 22 March). These insects were stored: • dry – Lepidoptera (moths), Curculionidae (weevils), Tephritidae (fruit flies); or • in absolute alcohol (guava moths only); or • in 70% alcohol – all Hymenopteran wasp specimens.

Photo: Examining the contents of a 2L container of incubated lilly pilly fruit.

DNA analysis of guava moth populations Ecogene (Landcare Research) sequenced 658 base pairs of the Cytochrome oxidase I (COI) gene from the following guava moth specimens:

• 10 guava moths from four of the lilly pilly fruit samples collected in and around Sydney on 20 and 22 March 2012

• 10 larvae of guava moth infesting feijoa in Mangonui and Cable Bay (Doubtless Bay, Northland) on 17 May 2012

• the leg of one pinned moth specimen from New Caledonia held at L’Institut agronomique néo-Calédonien

RESULTS AND DISCUSSION

Pheromone trapping of guava moth Very few guava moths have been caught in pheromone traps baited with synthetic guava moth sex pheromone in Australia. Only one moth was trapped in this study, in Bundaberg, QLD, in November 2011. It was identified as C. improbana using morphological characters (R. Hoare, pers. comm.). No guava moths were caught in a pheromone trap placed in a lilly pilly tree from 23–28 April 2012, in Stanmore, an inner-west suburb in Sydney. Fruit from this tree had yielded guava moth during sampling a month earlier. C. improbana has only previously been collected in guava moth sex pheromone trap in Nambour, QLD, in August 2004 (2 moths trapped in 4 days in a derelict guava orchard, Dymock 2004, Northland Regional Council Internal Report). All pheromone traps were baited with DESIRE 2-component synthetic guava moth sex pheromone. Insects reared from lilly pilly fruit Fruit was collected from lilly pilly from 28 sites in northern Sydney, Sydney north shore and inner-west suburbs on 20 and 22 March 2012 (data available as a separate excel file). C. improbana was reared from lilly pilly fruit from 23 of the 28 sites. A total of 680 C. improbana moths emerged from fruit samples during the study (Table 1).

Table 1. Insects reared from lilly pilly fruit collected from Sydney, Australia, in March 2012.

Insect reared Feeding niche No. of host samples (total samples = 28)

No. of specimens recorded

No. of samples where guava moth conjointly reared

Coscinoptycha improbana - guava moth

seed feeder/frugivore

23 680

Sigastus spp. weevils

seed feeder 15 64 13

Fruitfly - Tephritidae

frugivore 9 36 7

Other moths (Tortricidae)

various 7 8 6

Anselmella miltoni wasp

seed feeder 13 not recorded, >100

12

Megastigmus wasp

seed feeder 10 not recorded, >100

10

Braconidae wasps parasitoid 6 awaiting confirmation

6

Ichneumonidae wasps

parasitoid 2 awaiting confirmation

2

Parasitoid wasps from the Hymenopteran superfamily Ichneumonoidea (Families Ichneumonidae and Braconidae) were reared from eight of the 28 sites, representing at least five different species. All were reared from samples where C. improbana were infesting fruit. All Ichneuomonoidea wasp specimens were 7–8 mm in length. The wasp specimens from the family Ichneumonidae have been tentatively identified by Dr Andrew Austin, University of Adelaide, as being in the genus Eriborus. It is likely that the Ichneumonoidea wasps collected during this project are, as yet, undescribed species.

Photo: Specimens of Ichneumonoidea reared from Syzygium paniculatum collected from 16 McKenzie St, Lavender Bay, north of Sydney Harbour Bridge.

Other Hymenopteran wasps reared from the samples were Anselmella miltoni (Hymenoptera: Eulophidae) from 13 sites, and Megastigmus sp. (Hymenoptera: Torymidae), from ten sites. Both these species are known as seed feeders. Megastigmus sp. is thought to be associated commensally with A. miltoni (Juniper and Britton, 2010). These two species are too small (<2 mm in length) to be considered hosts of the Ichneumonoidea reared from the fruit samples.

Ichneumonoidea parasitoid wasps were recovered from mixed samples of insects that included a number of species not targeted for biocontrol in this project. These non-target insects are large enough to sustain the development of the Ichneumonoidea parasitoid wasps and therefore are potential parasitoid hosts. These were:

• the Sigastus weevils

• moths from the family Tortricidae

• fruitfly (Tephritidae).

Sigastus sp. weevils (Curculionidae) Two species of seed-feeding weevils, Sigastus spp., were reared from 15 sites (64 weevils reared). Larvae and adults, 4–5 mm in length, and are potential hosts of the Ichneumonoidea reared from the samples. Sigastus sp. weevils have been reared from macadamia on the Atherton Tableland, QLD, in 2002 (Fay et al. 2001). Ros Blanche at the former Rainforest Centre for Research found that, while 11 species of Hymenoptera wasp were reared from the weevils (some of the wasps are being considered as biocontrol agents of the weevil), none were from the Ichneumonidea (Blanche et al. 2002, Ros Blanche and Saul Cunningham, CSIRO Canberra, pers. comm.).

It is possible that the presence of these weevil larvae in fruit could deter guava moth from ovipositing on fruit and/or guava moth larvae successfully developing within fruit. There was no relationship between the number of Sigastus spp. weevils and the number of guava moths emerging from fruit samples in this study (R2=0.062, t=1.25, 23 df). At least 95% of fruit from S. paniculatum collected in Kerikeri in May 2012 and from Awanui in June 2011 were infested by an unnamed weevil larva. Attempts to rear this weevil for identification have so far proven unsuccessful. However, in New Zealand, this weevil could be excluding guava moth from the fruit as no guava moth larvae have been found either infesting S. paniculatum fruit or co-habiting with these weevil larvae inside the fruit. Other moths (Tortricidae) Eight leafroller moths (Lepidoptera: Tortricidae) in total were reared from seven sites. They are also potential non-target hosts of the Ichneumonoidea wasps found in this study. Fruit fly (Tephritidae) A total of 36 adult Tephritidae fruit flies (Bactrocera sp., probably jarvisi) were reared from fruit collected at nine of the lilly pilly collection sites. The presence of Tephritid larvae in fruit samples may deter guava moth oviposition or have a detrimental effect on guava moth larval survival within fruit, but again there was no relationship between the number of Tephritid flies and the number of guava moths emerging from fruit samples (R2 = 0.097, t=1.6, 23 df). No wasp specimens of the genus Diachasmimorpha (which parasitise Tephritidae in Australia), were reared from any of the samples. DNA analysis of guava moth populations Ecogene sequenced 658 base pairs of the Cytochrome oxidase I (COI) gene from 10 larval specimens of guava moth from Doubtless Bay, New Zealand, and 10 moth specimens from Sydney, Australia. Unfortunately, there was insufficient DNA material obtained from New Caledonian guava moth material (only a leg) to sequence DNA to determine the taxonomic status of New Caledonia guava moth species. Of the ten New Zealand specimens:

• nine were identical, with the same three base pairs different from the Sydney specimens

• one New Zealand specimen clustered with two Sydney specimens. Therefore, from the guava moth samples analysed, there are two haplotypes present in New Zealand, with one haplotype appearing unique with respect to Sydney specimens.

Key to collection sites

NZ – 10 larvae collected from feijoa in Cable Bay and Mangonui, Doubtless Bay, 17 May 2012 AUSMorgan – 2 moths that emerged from Syzygium australe hybrid – corner of Morgan St and Livingstone Rd, Marrickville, inner-west Sydney AUSChurch –3 moths that emerged from Syzygium paniculatum – corner of Pittwater and Baker Rds, Church Point, Northern Beaches 33 km NE Sydney CBD AUSBigola – 2 moths that emerged from Syzygium paniculatum – Bigola suburb, Northern Beaches, 33 km NE Sydney CBD AUSGrace – 3 moths that emerged from Syzygium paniculatum – 11 Grace Ave, Frenchs Forest, North Shore Sydney

Existing data for C. improbana analysed by Dr Andrew Mitchell, Australia Museum, Sydney, using the Barcode of Life Database (BOLD), showed very little variation in the COI gene in the 14 specimens sequenced so far, collected inland from Canberra to Orange, and Sydney to the QLD border on the coast.

Neither lilly pilly nor acmena fruit have been recorded as hosts of guava moth in Northland (Lisa Jamieson pers. comm. and pers. obs.), despite being readily available to guava moth. Fruit of macadamia (Ruth Huwer, pers. comm.), feijoa and loquat and have not been recorded as hosts of guava moth in Australia – including fruit from four feijoa trees in Sydney examined during this study (Table 2). The most common insect infesting macadamia nuts in Queensland is the macadamia nut borer, Cryptophlebia ombrodelta (Lepidoptera: Torticidae), which may be excluding guava moth from infesting macadamia nuts.

Table 2. Host fruit of guava moth common to Australia, New Zealand and Norfolk Island + host - present, but not a host ? present, but not known if a host

Species Plant Family

Place of Origin

Australia New Zealand

Norfolk Island

Guava, Psidium sp.

Myrtaceae Sth America + + +

Feijoa, Acca sellowiana

Myrtaceae Sth America - + Not present

Lilly pilly – Syzygium paniculatum, S. australe

Myrtaceae Australia + - Not present

Acmena smithii

Mytaceae Australia + - Not present

Macadamia spp.

Proteaceae Australia - + ?

Citrus spp. Rutaceae SE Asia + + ? Loquat, Eriobotrya japonica

Rosaceae China - + ?

Peach, Pyrus sp.

Rosaceae China - + +

Plum (Old World), Prunus spp.

Rosaceae Europe - + ?

Pear, Pyrus pyrifolia

Rosaceae Europe - + ?

It is not known whether the differences in host fruit choice by guava moth in New Zealand, Australia, Norfolk Island and New Caledonia guava moths is due to the:

• presence of different biotypes of guava moth with different feeding preferences, or

• presence of different insects co-habiting or competing for fruit, or

• differing host fruit range available. Further data are required from guava moths across a broader geographical range (including New Caledonia and Norfolk Island) to determine the origin of the New Zealand guava moth haplotypes. In the meantime, the differences recorded thus far are not great enough that any NZ specimens can be recognised as representing separate species. At present, C. improbana, as a single species, is still considered as the target for a biocontrol programme.

FUTURE PLANS

The small amount of variation seen in the COI gene is typical of within-species differences. Therefore, guava moths identified by morphological characteristics in New Zealand and Australia can still be considered a single species. The present data suggest that the geographic source of the NZ guava moth is mixed and includes areas other than where guava moth was collected in Sydney. More samples from different regions in Australia and NZ would be needed to confirm this. Collections of guava moth in New Caledonia and Norfolk Island are planned.

Confirmation that the Ichneumonidea parasitoid wasps reared during this project are indeed parasitizing guava moth is the next step in a biocontrol programme for guava moth. The wasp species identified as potential biocontrol agents for guava moth would be exposed to guava moth-infested fruit of lilly pilly, in the absence of other insect species, to determine the level of parasitism.

REFERENCES

Blanche, R., Bauer, R., Cunningham, S. and Floyd, R. (2002). Services and dis-services of rainforest insects to crops in north Queensland. Cooperative Research Centre for Tropical Rainforest Ecology and Management. Cairns. (20 pages)

Common, I. F. B. (1990). Moths of Australia. Melbourne University Press, Carlton, Australia

Fay, H. A. C., De Faveri, S. G., Storey, R. I. and Watson, J. (2001) Sigastus weevil – an emerging pest of macadamias in north Queensland. Proceedings of the 6th Workshop in Tropical Agricultural Entomology, Darwin, NT, 11–15 May 1998. NT-DPIF Technical Bulletin No. 288. Gibb, A. R., Suckling, D. M., Morris, B. D., Dawson, T. E., Bunn, B., Comeskey, D. and Dymock, J. J. (2006). (Z)-7-Tricosene and Monounsaturated Ketones as Sex Pheromone Components of the Australian Guava Moth Coscinoptycha improbana: Identification, Field Trapping, and Phenology . Journal of Chemical Ecology 32(1): 221–237. Jamieson, L. E., Dymock, J. J., Dawson, T., Froud, K. J., Seldon, D. S., Suckling, D. M., and Gibb, A. R. (2004). Guava moth in New Zealand – Distribution, hosts, lifecycle observations and discussion of pest management options. New Zealand Plant Protection 57: 13–19.

Juniper, P.A. and Britton, D.R. (2010). Insects associated with the fruit of Syzygium paniculatum (magenta lilly pilly) and Syzygium australe (brush cherry). Australian Journal of Entomology 49: 296–303.

Mille, C., Munzinger, J. and Jourdan, H. (2012). First record of the Australian guava moth Coscinoptycha improbana Meyrick (Lepidoptera: Carposinidae) in New Caledonia: Implication for quarantine and biosecurity surveys in insular territories. Journal of Asia-Pacific Entomology , 15 (2) : 283–285. ACKNOWLEDGMENTS Many thanks to following people who assisted with this project Dr Andy Austin, University of Adelaide, identification of Ichneumonoidea Dr David Britton – Australian Museum, Sydney – making specimens from Australia Museum available for comparison Leonie Clunie – NZAC Landcare Research, Auckland, transporting guava moth specimens to Ecogene Dr Dianne Gleeson, Robyn Howitt, Dr Frank Molinia, Dr Celia Pruden – Ecogene, Landcare Research, Auckland – gene sequencing of guava moth Dr Robert Hoare – Landcare Research, Auckland – identifying guava moth from pheromone trap Dr Marianne Horak – checking guava moth specimens in the Australian National Insect Collection, Canberra Dr Ruth Huwer – NSW Centre for Tropical Horticultural Research, Alstonville – pheromone trapping for guava moth Dr Lisa Jamieson – Plant and Food Research, helpful advice during the project Dr Peter Juniper – navigating the suburbs of Sydney and collection of lilly pilly fruit samples Dr Iain Kay – DEEDI Bundaberg – pheromone trapping Dr Sarah Mansfield – University of Sydney, providing laboratory space for lilly pilly samples, laboratory facilities and equipment, transporting guava moth specimens to the Australian Museum and arranging for Ichneumonoidea specimen transport to Adelaide Dr Christian Mille – L’Institut agronomique néo-Calédonien – providing guava moth leg specimen for gene sequencing Dr Andrew Mitchell – Australian Museum, Sydney, DNA analysis of specimens of Carposinidae Dr Leigh Pilkington – DPI, Gosford, NSW – pheromone trapping Derek Smith – Australian Museum, arranging transportation of guava moth specimens to Auckland Wayne’s World Timber & Building Supplies, Botany Rd, Sydney – providing untreated sawdust