EPA report

Import and release of glorifica July 2013

Advice to the Decision Making Committee on application APP201710: – To import and release Limenitis glorifica (Honshu white admiral ), as a biocontrol agent for Japanese , under section 34 of the Hazardous Substances and New Organisms Act 1996

www.epa.govt.nz 2

EPA advice APP201710

Executive Summary and Recommendation

In May 2013, the Greater Wellington Regional Council made an application to the Environmental Protection Authority (EPA) seeking to import and release Limenitis glorifica as a potential biocontrol agent for Japanese honeysuckle ().

Host range testing shows that no native, valued and/or taonga will be adversely affected by this agent, and we recommend that it be approved for release.

. June 2013 3

EPA advice APP201710

Table of Contents

Executive Summary and Recommendation ...... 2

Table of Contents ...... 3

1. The application process ...... 4 Purpose of this document ...... 4 Submission process ...... 4 Submissions ...... 4 Application summary ...... 5 Background ...... 5

2. The organism proposed for release ...... 6

3. Host-range testing ...... 7

4. Minimum standards...... 11

5. The ability to establish an undesirable self-sustaining population and the ease of eradication ...... 14

6. Effects of any inseparable organism...... 14

7. Adverse and Positive effects ...... 14

References ...... 20

Appendix 1 ...... 21

Appendix 2 ...... 23

Appendix 3 ...... 24

. June 2013 4

EPA advice APP201710

1. The application process

Purpose of this document

1.1. This document has been prepared by We; Asela Atapattu (Manager, New Organisms), Kate Bromfield (Senior Advisor, New Organisms) and Manu Graham (Senior Advisor, Māori Policy and Operations), to advise the HSNO Decision Making Committee on the results of our risk assessment of an application to import and release Limenitis glorifica or Honshu white admiral butterfly, as a biocontrol agent for Japanese honeysuckle. The document discusses information provided in the application and other readily available sources. Submission process

1.2. The Greater Wellington Regional Council (GWRC) lodged an application with the EPA on 3 May 2013 to import and release Limenitis glorifica under section 34 of the Hazardous Substances and New Organisms (HSNO) Act (the Act).

1.3. Application APP201710 was publicly notified as required by section 53(1)(b) of the Act. The 30 working day notification period began on 6 May 2013 and closed on 18 June 2013.

1.4. Submitters were asked to provide information, make comments and raise issues, particularly with regard to, but not limited to the following matters:  methodology of the host-range testing;  adverse effects1, especially adverse effects not identified in the application, and  positive effects2, especially positive effects not identified in the application.

Submissions

1.5. Eleven submissions were received during the submission period in response to public notification of the application. Seven submitters; Hawkes Bay Regional Council, Northland Regional Council, Te Rūnanga o Ngāi Tahu, Rob Morton, Bay of Plenty Regional Council, Waikato Regional Council and Thames Coromandel Regional Council supported the application to import and release Limenitis glorifica. Two submitters; Donovan Scientific Research, and Kiwifruit Vine Health neither supported nor opposed the application. Two submitters; Huakina Development Trust and Cliff Mason opposed the application. One late submission, which neither supported nor opposed the application, was received on 8 July 2013, from the Moths and of New Zealand Trust

1 Adverse effects can include any risks and costs associated with approving the release of these organisms.

2 Positive effects can include any benefits associated with approving the release of these organisms.

. June 2013 5

EPA advice APP201710

(MBNZT). Cliff Mason and Jacqui Knight from MBNZT have both expressed an interest in speaking at a hearing. The submissions are summarised in Appendix 1.

Submissions from MPI and DOC

1.6. As required by the Act and the Hazardous Substances and New Organisms (Methodology) Order 1998 (the Methodology), the Ministry for Primary Industries (MPI) and the Department of Conservation (DOC) were advised of the application and provided with the opportunity to comment. We gave particular regard to the comments provided by DOC, and these full comments are provided in Appendix 2. MPI provided no comment on the application. Application summary

1.7. GWRC makes this application on behalf of the National Biocontrol Collective, a collective comprising 13 regional councils and the Department of Conservation (DOC). Landcare Research was the science provider for the application, and Richard Hill & Associates prepared the application and managed the application process on behalf of GWRC.

1.8. The application seeks approval to import and release the white admiral butterfly, Limenitis glorifica as a biocontrol agent for Japanese honeysuckle (Lonicera japonica).

1.9. GWRC consider that they can no longer manage Japanese honeysuckle using conventional means. The number of infested sites is growing, many sites are too remote to be feasibly accessible, and off- target damage to underlying vegetation from the use of herbicides over wide areas is unacceptable. GWRC consider biological control to be an appropriate tactic to use against this weed because biological control agents are self-dispersing and can locate isolated host plants.

1.10. The application presents evidence from field observations in Japan and from laboratory experiments in Japan and New Zealand, indicating that significant adverse effects are unlikely. Background

1.11. The biological control program against Japanese honeysuckle (Lonicera japonica) aims to limit the shading effects of the weed by reducing its biomass, slowing its vegetative spread, and reducing seed production. Japanese honeysuckle is a perennial vine that spreads by rhizomes, above-ground runners, and sometimes by seeds. It can infest forests but has the most impact on forest margins. It may form a complete blanket over small trees and shrubs in coastal situations, and is increasingly troublesome in shrub land, forest margins and open roadsides.

1.12. Japanese honeysuckle can out-compete many trees and shrubs, and native seedlings are unlikely to establish beneath it. This leads to a simplified vegetation structure with lower local biodiversity (Williams and Timmins 1998). Japanese honeysuckle is distributed from Northland to Stewart Island.

. June 2013 6

EPA advice APP201710

Williams and Timmins (1998) listed it as “widespread” or “spreading” in all North Island conservancies except Northland but less common in the South Island. While it is present almost throughout New Zealand, it is expanding in many areas, and its full distribution and impact on public conservation land is yet to become apparent. It occurs on public and private land, and is still relatively common as a cultivated in older gardens despite its Unwanted Organism status and being listed on the National Pest Plant Accord (NPPA).

1.13. Current management is limited to physical and chemical methods, often in combination. Physical methods involve cutting stems and detaching honeysuckle plants that are smothering native canopy . This is labor-intensive work, but has the advantage of reducing the need for herbicides and reducing herbicide damage to host trees and shrubs. Cut stems that remain in contact with the ground are then sprayed or swabbed with herbicide. Chemical methods alone usually involve spraying, either from the ground or helicopter, depending on the scale of the control operation and available resources, although accessibility in remote areas is also a factor.

1.14. Japanese honeysuckle was first recorded as being naturalised in New Zealand in 1926, but is known to have been cultivated here as an ornamental since 1872 (Williams and Timmins 1998). It is a hardy plant, tolerant of cold winter temperatures and a wide range of soil types, including poorly draining soils and those high in salts or heavy metals (Williams et al. 2001). As well as Japanese honeysuckle, three other Lonicera spp. are naturalised in New Zealand: Lonicera × americana; L. nitida; and L. periclymenum.

2. The organism proposed for release

2.1. This organism has the following :

Class: Insecta

Order:

Family:

Subfamily:

Tribe:

Genus: Limenitis

Species: glorifica Fruhstorfer, 1909

2.2. There are two closely related Limenitis species that feed on honeysuckle in Japan. (Eurasian white admiral) is widely distributed in Japan from Hokkaido in the north to Kyushu in the south (Tanaka 1978). It prefers forest margins of cold-temperate areas throughout Eurasia, and field

. June 2013 7

EPA advice APP201710

experiments and experimental studies have shown that larvae can feed on plants from a number of subgenera and sections of the Lonicera, as well as Weigela3.

2.3. Limenitis glorifica (Honshu white admiral) is endemic to Honshu, one island of Japan. It is found in the open, or in light shrub land in comparatively dry warm-temperate habitats. Unlike Limenitis camilla, records suggest it is restricted to several species of Lonicera. In Honshu it is widely distributed from the western lowlands of Yamaguchi Prefecture to Shimokita Peninsula in the north. It can be found to about 1000 m above sea level but is most common in the lowlands below. The altitudinal distribution limit is 1450 m (Tanaka 1978).

2.4. In subtropical warm central Honshu, L. glorifica completes three to four generations each year, but in cold mountainous zones there is typically one generation annually. Adult butterflies emerge from their chrysalis in spring. Mating is behaviourally complex, and may occur on the wing high in the air. Eggs are laid on the foliage of Lonicera japonica. Larvae spend winter inside shelters constructed from honeysuckle leaves and hanging from the stem. Pupation is induced in late summer once daylight falls below about 13–14 hours (Tanaka 1978).

3. Host-range testing

3.1. The applicant has demonstrated that L. glorifica poses no threat to native flora through off-target effects. We consider that L. glorifica is host specific, and will only act as a biocontrol agent for Japanese honeysuckle.

3.2. Wapshere (1974) states that the plants most likely to be attacked by a proposed biocontrol agent are those most closely related to the target weed and growing in the area where that agent will be used. He developed the centrifugal phylogenetic testing system for selecting test plants, which accurately predicts the range of hosts that the control agent will use in its new environment. This system has been universally adopted by biological control practitioners worldwide and continues to be effective at preventing unexpected attacks of non-target plants in the field. The technique was revised by Briese (2005), reducing the need for distantly related “safeguard” plants to be included in test lists, as these have been found to add no additional value to host range testing procedures.

3.3. The applicant has provided evidence that while L. camilla has a relatively broad host range (feeding on several Lonicera and species), L. glorifica is feeds almost exclusively on Lonicera japonica. Tanaka (1978) reported finding one L. glorifica larva on Lonicera morrowii. At

3 Weigela is a genus of potentially 38 species of deciduous shrubs in the family .

. June 2013 8

EPA advice APP201710

another site, adult L. glorifica were found flying in the vicinity of Lonicera morrowii, but no eggs were found. Tanaka reported that, on one occasion, a female L. glorifica was observed to lay a single egg on a Weigela floribunda plant. However, although the butterfly chose to lay its egg there, the resultant larva did not feed on the foliage. He noted that the Weigela floribunda plant in question was growing alongside Lonicera japonica, so it is possible that oviposition on Weigela floribunda was stimulated by the proximity of Lonicera japonica4.

3.4. Host-range tests for L. glorifica were performed by Tanaka (1978), using eggs or first-instar larvae collected from L. japonica and transferred to test plants or L. japonica controls. He found that larvae fed well on L. japonica, L. morrowii, and L. gracilipes, but noted that when feeding on L. morrowii larval growth was retarded (compared to L. japonica), and the resultant adults were small or expanded their wings imperfectly on emergence. He suggested that this indicated that L. morrowii is a sub- optimal host. Larvae did not mature to adult on L. gracilipes, but this may have been due to declining host-plant quality in autumn (Tanaka 1978). Generally, larvae did not feed at all on Weigela floribunda, and for the few exceptions, the developmental period was prolonged and the single adult reared was very small-sized. Larvae did not feed on Abelia spathulata at all.

3.5. A selection of host-range tests were conducted by Landcare Research staff at Tsukuba, Japan in 2012. The results are summarised in Table 1, and it is valuable to note that none of the species listed are New Zealand natives.

3.6. Almost all L. glorifica larvae transferred to Japanese honeysuckle foliage fed normally and survived to emerge as adults (81%). However, one larva (4.5%) survived on Weigela formosa, which belongs to the subfamily Diervilloideae, so the physiological host range for larval development may also include this subfamily. Within the subfamily Caprifolioideae, the evergreen Lonicera nitida could not support larval development at all. This indicates that the physiological host-range of L. glorifica does not include all Lonicera species. However, the host range was not restricted solely to the Lonicera genus, as 26.1% of larvae survived to adulthood on Leycesteria formosa.

3.7. The applicant states that to be ecologically fit, an individual must not only complete its developmental life cycle, but also be long-lived and fecund. Larvae reared on Symphoricarpus sp. and Heptacodes micanoides took longer to complete development than those fed on Lonicera japonica and Leycesteria formosa, and produced smaller pupae. Low pupal weight is often an indicator of how many eggs a female will lay in her lifetime. Similarly, larvae fed on Lonicera japonica and

4 An insect’s acceptance of a non-target species can temporarily increase as a result of being stimulated by the chemistry of nearby target plants (Marohasy 1998).

. June 2013 9

EPA advice APP201710

Leycesteria formosa produced adult butterflies that survived for approximately 10 days, but adults produced on other plants died almost immediately. Thus, although the larvae may be able to complete development on a few non-native, ornamental plants other than Lonicera japonica (physiological host- range), the species does not thrive on them, and the actual host-range is likely to be narrower. Lack of survival on Viburnum sp., which belongs to the family Adoxaceae, reinforced this view.

Table 1 Survival and development of first-instar Honshu white admiral larvae transferred to test plant foliage in tests conducted by Landcare Research staff at Tsukuba, Japan, in 2012 (see Landcare Research (2013c) for testing details; no native plants required testing; - = no test;).

Family Subfamily Test plant species PLS5 PLSP6 Tn7 % survival to adult

Caprifoliaceae Caprifolioideae Lonicera japonica 18/20 9/10 20(10)* 81

Lonicera nitida 0/20 - 20 0

Symphoricarpus sp. 10/20 1/10 20 5

Leycesteria formosa 7/20 6/10 23 26.1

Heptacodium micanoides - 1/10 10 10

Diervilloideae Weigela formosa 8/20 1/10 22 4.5

Linnaeoideae Abelia sp. 0/20 - 20 0

Morinoideae Morina longifolia 0/20 - 20 0

Dipsacoideae Dipsacus sylvestris 0/20 - 20 0

Valerianoideae officinalis 0/20 - 20 0

Adoxaceae Viburnum sp. 0/20 - 20 0

* 20 larvae were used to measure survival of first instar larvae; 10 of the 18 surviving larvae were used in the development test.

3.8. The applicant considers that the evidence from their host range testing indicates that:  The fundamental or physiological host range of L. glorifica is confined within the family Caprifoliaceae;  There are no native species in this family so there is no risk to New Zealand’s native flora;  Larvae placed on Weigela (subfamily Diervilloideae) mostly died, but of the 4.5% that survived, development time was extended, pupal weight was low and adult survival was short, indicating that it is unlikely that this plant could support a self-sustaining population of L. glorifica;

5 Proportion of larvae surviving 10 days in petri dishes

6 Proportion of larvae surviving to adult in potted plant tests

7 Total number of larvae used in tests

. June 2013 10

EPA advice APP201710

 Limenitis glorifica will have a narrow host range in New Zealand, confined to species in the subfamily Caprifolioideae;  Not all species in this subfamily will be hosts for L. glorifica as three of the five plants tested did not provide adequate nutrition for L. glorifica larvae;  Moderate survival on Leycesteria formosa indicates this weedy species may be a suitable host for L. glorifica in New Zealand;  Technical difficulties preclude conducting oviposition tests to better define the realisable host range of this insect. It is likely that some plants capable of supporting larval development will not be selected as oviposition sites by the butterfly, narrowing the host range further; and  The potential host range in New Zealand will be narrow, but some ornamental Lonicera spp. (e.g. L. periclymenum) may be suitable host plants.

3.9. The applicant had their assessment of the host range of L. glorifica peer reviewed by an independent expert in the field, Dr Toni Withers, forest entomologist from Scion, specialising in biological control. Her comments are provided in the application and summarised here. The applicant provided EPA staff with a response to Dr Withers review, and this is included in Appendix 3 of this document.

3.10. Dr Withers noted that the “significant issue here is that New Zealand has no native in the family Caprifoliaceae. Therefore none of the plants tested are native or considered to be particularly valued ornamentals in New Zealand. Did Landcare Research consider whether some other closely related native plants should be tested as an outgroup, even if they do not share the form of a “honeysuckle” vine?”

3.11. The applicant responded that they “did consider including native plant species during the development of the test plant list. However, the inclusion of native species was considered unnecessary because Japanese honeysuckle is so taxonomically remote from the most closely-related native species”.

3.12. Dr Withers had concerns over the statistical significance of the host range tests provided by the applicant (that the applicant has addressed in their response), but she noted that “in light of the lack of native Caprifoliaceae, [the applicant’s] interpretations of the risk posed by L. glorifica to these weedy and ornamental species appear to be conservative and entirely appropriate to the data that I have scrutinised.”

3.13. We consider Dr Withers an appropriate independent reviewer, and we are satisfied with the response provide by the applicant. In addition, Te Rūnanga o Ngāi Tahu noted in their submission: “No native species were included in the host-specificity testing, a drawback which would normally cause us considerable concern. However, given the very considerable taxonomic distance between L. japonica and the nearest native New Zealand plant, we do not see this as precluding the introduction of the agent”.

. June 2013 11

EPA advice APP201710

3.14. Barry O’Neil, on behalf of Kiwi Vine Health commented in his submission that even if “the Wapshere methodology was being fully applied and as such we could have every confidence that while no containment host specificity trials are being undertaken on commercial horticultural species, they wouldn’t be a host species for the agent….It is very difficult to explain to lay people, especially when the application focuses on risks to native plant species…, with hardly any mention of the risks to the commercial horticultural sector and no containment testing having been undertaken on commercial plant species”.

3.15. Cliff Mason submitted that “the host testing programme is inadequate, especially as it includes no native species” He went on to say “the theoretical basis of the Applicant's methods and conclusions are based on taxonomy that is artefactual and unstable”, and that “the proposed introduction is an experiment conducted upon the New Zealand environment”.

3.16. However, Waikato Regional Council submitted that “Landcare Research and their associates have a long history of rigorous testing and successful biocontrol releases. The testing methodology appears rigorous and in keeping with international best practice”.

3.17. We consider that centrifugal host testing, as developed by Wapshere (1974) and Briese (2005), and employed by the applicant in their host range testing, represents best practice with regards to test species, and we recognise that Landcare Research has an international reputation in this field. The applicant’s methods and conclusions presented in the application are robust and represent a fair assessment of the likely effects of the proposed introduction. The introduction of L. glorifica is very unlikely to have any adverse effects on plants that are not native to New Zealand, but are still considered valuable, because host range testing indicates that plants other than Lonicera japonica are sub-optimal for L. glorifica development. Limenitis glorifica will have a narrow host range in New Zealand, confined to species in the subfamily Caprifolioideae. We therefore consider that there is no scientific uncertainty around the potential host range of L. glorifica.

4. Minimum standards

4.1. Prior to approving any new organism the EPA is required to ensure that if the organism were to be released, it would meet the following minimum standards set out in section 36 of Act.

Section 36 (a): whether Limenitis glorifica is likely to cause any significant displacement of any native species within its natural habitat.

4.2. The applicant has provided information to the effect that only three parasitoids of nymphalid butterflies are known in New Zealand. Barron et al. (2003) recorded parasitism of red admirals by native ‘Telenomus’ sp., and recorded highly variable pupal parasitism by the introduced parasitic wasps

. June 2013 12

EPA advice APP201710

Echthromorpha intricatoria (Ichneumonidae) and Pteromalus puparum (Pteromalidae). Although red and yellow admirals belong to a different subfamily from the Honshu white admiral, their life stages have similar habits, and Honshu white admiral may be susceptible to the generalist, exotic pupa parasitoids. Paynter et al. (2010) found that control agents introduced to New Zealand could attract native parasitoids if there was a related native congener8 that fed in an ‘ecologically analogous’ habitat. While L. glorifica feeds in a similar way to the red admiral (on foliage), the wide taxonomic gap between their host plants will likely limit the host-finding ability of ‘Telenomus’ sp., the one native parasitoid known to attack the red admiral.

4.3. We therefore consider that it is extremely unlikely that L. glorifica will cause displacement of any native butterflies in their native habitat.

4.4. We consider that for L. glorifica to cause displacement of native parasitoids through spill over from native congeners is unlikely, and if it were to occur, would not have a significant impact of the life history of the native parasitoid.

Section 36 (b): whether Limenitis glorifica is likely to cause any significant deterioration of natural habitats.

4.5. Host testing indicates that no native plants will be affected by the introduction of L. glorifica. After considering the available information, we consider that it is extememly unlikely that the release of L. gloriffica could cause significant deterioration of native habitats.

Section 36 (c): whether Limenitis glorifica is likely to cause any significant adverse effects on human health and safety

4.6. We have searched the literature, and there are no examples of L. glorifica acting as a human pathogen or posing a threat of any kind to human safety. We have not identified any effects on human health and safety and therefore consider that L. glorifica is not likely to cause significant adverse effects on human health and safety.

Section 36 (d): whether Limenitis glorifica is likely to cause any significant adverse effect to New Zealand’s inherent genetic diversity

4.7. The applicant has provided evidence that L. glorifica does not have a close taxonomic affinity with any New Zealand species. There are about around 17 butterfly species resident on the New Zealand mainland (http://nzbutterfly.info/). Nine of these 17 species belong to the Nymphalidae (the same family as L. glorifica), including red and yellow admirals, and Monarch butterflies. Red and yellow admirals belong to the subfamily and feed exclusively on native or introduced nettles.

8 a person, organism, or thing resembling another in nature or action

. June 2013 13

EPA advice APP201710

Monarch butterflies belong to the Danainae, and are restricted to swan plant and one or two related plants. The remaining six species are native species belonging to the subfamily Satyrinae, and their larvae feed on native grasses. None of these species present in New Zealand belong to the subfamily Limenitidinae, and therefore, they are biologically incapable of breeding with L. glorifica. Significant populations of Honshu white admiral larvae will only be found in association with Japanese honeysuckle. No New Zealand nymphalids, native or introduced, feed on this plant.

4.8. Other butterfly species on mainland New Zealand are even less closely related to L. glorifica. Seven belong to the family Lycaenidae. Three of these (the blue butterflies) feed on native and exotic legumes. Four species (the copper butterflies) feed on Muehlenbeckia spp. - native vines that are sometimes found associated with Japanese honeysuckle. The cabbage white butterfly (family Pieridae), is also present in New Zealand, but is an exotic pest that feeds only on brassicas. These species are biologically incapable of inter-breeding with L. glorifica.

4.9. The Huakina Development Trust made a submission to the effect that “we do not know the potential harm the butterfly may cause [so] it is very difficult to say where any impacts could occur”. They consider all native species of the whenua to be taonga.

4.10. We acknowledge that the introduction of any new organism to New Zealand has the potential to cause harm to New Zealand’s genetic diversity. However, the biology and the taxonomic classification of L. glorifica indicates that this potential is highly unlikely to eventuate as it cannot cross-breed with any species present in New Zealand. We therefore consider that L. glorifica is unlikely to cause any significant adverse effects to New Zealand’s inherent genetic diversity.

Section 36 (e): whether Limenitis glorifica is likely to cause disease, be parasitic, or become a vector for human, , or plant disease.

4.11. We have searched the literature, and there are no examples of L. glorifica acting as a pathogen, parasite, or vector of human, animal, or plant disease. We therefore consider that L. glorifica is not likely to cause disease, be parasitic, or become a vector for human, animal, or plant disease.

Conclusion on the minimum standards

4.12. We considers that L. glorifica is unlikely to cause significant displacement of other organisms, cause significant deterioration of natural habitats, have any significant adverse effects on human health and safety, or have significant adverse effects on New Zealand’s inherent genetic diversity. It is unlikely to cause disease, be parasitic, or become a vector for human, animal, or plant disease.

4.13. Therefore, we consider that Limenitis glorifica meets the minimum standards as stated in the Act.

. June 2013 14

EPA advice APP201710

5. The ability to establish an undesirable self-sustaining population and the ease of eradication

5.1. Section 37 of the Act requires EPA staff to have regard to the ability of the organism to establish an undesirable self-sustaining population and the ease with which the organism could be eradicated if it established such a population. We consider that the purpose of introducing a biocontrol agent is to establish a self-sustaining population; however, this population would not be undesirable.

5.2. The Huakina Development Trust specifically asked in their submission “what is the ease with which the organism could be eradicated if it should start to become a pest?”

5.3. We note that the eradication of such a population would be very difficult, but that this is unlikely to be an objective. Host range testing indicates that populations of L. glorifica will not damage native plants, and the taxonomic classification indicates that it will not compete with native . We therefore consider that L. glorifica is unlikely to become a pest and require eradication.

6. Effects of any inseparable organism

6.1. There are no known inseparable organisms associated with Limenitis glorifica.

7. Adverse and Positive effects

7.1. Biological control agents can take many years to establish widely and have an impact on the target species. There is some uncertainty about whether L. glorifica will establish and disperse successfully, and how long this will take. If it does not establish, it can be assumed that there will be no effect (adverse or beneficial) from the release. Conversely, if it successfully establishes, effects may be significant. Therefore, in conducting this risk assessment we have assumed that L. glorifica will become widely established. Adverse effects

7.2. The applicant has identified potential adverse effects associated with the release of L. glorifica in addition to those discussed in section 4 of this document (Minimum Standards). In particular, the applicant identifies possible impacts of L. glorifica on the New Zealand environment if:  A decline in Japanese honeysuckle abundance leads to invasion of sensitive habitats by worse weeds; or  Reduced Japanese honeysuckle fruit production significantly limits the food supply of native birds.

. June 2013 15

EPA advice APP201710

7.3. Barry Donovan wrote in his submission on behalf of Donovan Scientific Insect Research, that “the possibility of adverse impacts on bumble bees and the potential for negative consequences has not been addressed”. He went on to say that unpublished data suggests that bumblebees are moderately attracted to Lonicera sp. and the plant may be considered an important foraging plant for bumblebees.

7.4. We consider that the introduction of L. glorifica will not cause a rapid decrease in the existing populations of Lonicera japonica, and that bumblebees have the opportunity to access numerous other foraging plants that are equally attractive to bees.

7.5. After considering the available information, we consider that the adverse effects associated with the release of Limenitis glorifica are negligible. Positive effects

7.6. The applicant has identified potential positive effects on the environment, on society and communities, and on the market economy, associated with the release of L. glorifica. They consider that the release of L. glorifica has the potential to:  reduce shading and increase the fitness of native plants growing beneath Japanese honeysuckle canopies;  reduce shading by Japanese honeysuckle and increase establishment of native seedlings in forest margins;  reduce growth of the weed and limit Japanese honeysuckle density in existing sites;  reduce fruit production and the rate of spread of Japanese honeysuckle to new sites;  benefit native food webs and other ecosystem functions through reduced shading and the recovery of native plants; and  reduce herbicide use.

7.7. Cliff Mason submitted that “the introduction of yet another alien species would further damage the ecology of New Zealand”. He considers that “the principal vector in the dispersal of L. japonica is human activity and changing this vector is preferable to changing the ecology of our country”. From a personal perspective, he states that “I do not wish the experience of nature in this country to be even more completely composed of encounters with introduced species”. However, Jacqui Knight submitted on behalf of MBNZT that “the introduction of another butterfly species would add to the pleasure that New Zealanders receive when they are in the outdoors”.

7.8. Waikato Regional Council submitted that “Japanese honeysuckle is a very serious weed and a true intractable species in the Waikato”. They note that “it is widespread and council’s biosecurity staff believe it has not reached its limits or true potential. The Waikato’s diverse climate makes this region ideal for the spread and establishment of Japanese honeysuckle”.

. June 2013 16

EPA advice APP201710

7.9. Cliff Mason submitted that “many of the places in which L. japonica is perceived as a pest are the same marginal habitats in which a multitude of pest plant species are potentially or actually a problem. Control of one pest species is unlikely to be of ecological benefit”.

7.10. We agree that there are multiple pest plants that councils and DOC contend with, and note that Waikato Regional Council commented that “managing this plant is very difficult and requires constant vigilance. Repeated control of regrowth is critical to managing the spread and eradication. Herbicides are the preferred means of control as the cut stump will regrow. Biocontrol has the potential to help reduce the use and dependence of herbicides, and reduce the invasiveness of Japanese honeysuckle”.

7.11. Darin Underhill, from the Hawke’s Bay Regional Council, commented in their submission that L. japonica is “well established in Hawke’s Bay and is seriously affecting native bush in some areas”. He considers that it is “at a stage where it can no longer be controlled by conventional means…because there are too many sites, a large proportion of these sites are inaccessible and damage to non-target vegetation would be detrimental”.

7.12. Jenny Dymock wrote in her submission on behalf of the Northland Regional Council that “while Japanese honeysuckle is not a widespread weed in Northland, it has the potential to disperse from existing localities…” She notes that “Japanese honeysuckle, together with Himalayan honeysuckle Leycesteria formosa, also found to be a minor host of Limenitis glorifica, is listed in the Northland Pest Management Strategy as a Community Pest Control Area pest plant and [both] are banned from sale and distribution in Northland”.

7.13. Rob Morton submitted on how hard Japanese honeysuckle is to control. He states that “hand clearance….is very difficult” and considers that “a biological control would be wonderful and give native bush restorers in the North of New Zealand a great boost in their efforts”.

7.14. Te Rūnanga o Ngāi Tahu submitted that they support this application “because of the potential of the agent to have a beneficial impact on infestations of the weed, and hence reduce the weed’s damaging effects on native species”. While they note that it is the number one weed in Marlborough, they commented that it is not a top priority weed in Te Wai Pounamu. However, they commented that “the widespread, localised infestations are serious enough to be actively controlled by spraying and/or physical methods (cutting stems)”, and “It seems likely that the weed will spread more widely, particularly with any increase in overall temperature due to climate change”.

7.15. We consider that as control of Japanese honeysuckle is effected by L. glorifica, resources currently being allocated to its control could be diverted into control of other pest weeds species. The long term aim of the application is for habitat restoration for the enjoyment of the wider community, and we consider that the introduction of L. glorifica can help achieve this aim. . June 2013 17

EPA advice APP201710

7.16. Having evaluated the information, we consider that there are long term environmental benefits that can be accredited to the release of Lonicera glorifica, and that these benefits are non-negligible. Effects on Māori and their culture and traditions and the principles of the Treaty of Waitangi (Te Tiriti o Waitangi)

Relationship of Māori to the environment 7.17. The potential effects of the import and release of L. glorifica on the relationship of Māori to the environment have been assessed in accordance sections 6(d) and 8 of the HSNO Act. Under these sections all persons exercising functions, powers, and duties under the Act shall take into account the relationship of Māori and their culture and traditions with their ancestral lands, water, taonga and the principles of the Treaty of Waitangi (te Tiriti o Waitangi).

7.18. In consideration of these functions and duties, this section of the report provides an overall evaluation of the consultation process with Māori that was undertaken by the applicant and their response to issues that were raised from this. Finally an assessment of the impact this application may have on the principles of te Tiriti o Waitangi has been provided.

Consultation and Assessment 7.19. The EPA policy on consultation with Māori requires that consultation be undertaken by the applicant in the first instance, and should lead to the effective exchange of information between the applicant and iwi/Māori as appropriate. In addition, another purpose of consultation in this context is to lead to the provision of information to the decision makers to enable them to evaluate risks, costs and benefits and make informed decisions in accordance with their legal duty under the Act.

7.20. To fulfil this requirement the applicant sent out over 200 letters and emails to members of the EPA Māori National Network which outlined the purpose of the application, proposed research they would undertake and made an invitation for parties to make comments/queries.

7.21. Several respondents supported the proposal. None opposed it, but several questions were raised about the following issues:

 Concerns about the threat to native species (biophysical and cultural);

 The impact to native or valued ecosystems; and

 The effectiveness of the agent and whether alternative methods of control were more viable.

7.22. The applicant responded to each concern raised both directly with the individuals raising them, and in their application. In summary their responses confirmed that risks to native species and ecosystems would be minimal and that the pest plant was itself a cause for significant concern to native environments. The applicant considered the release of the agent would support the protection of native species and ecosystems.

. June 2013 18

EPA advice APP201710

7.23. Ngā Kaihautū Tikanga Taiao (the EPA’s Māori advisory committee) reviewed the application. They advised that they are concerned at the push for more biological controls to be introduced across New Zealand.

7.24. Ngā Kaihautū acknowledges that:

 Japanese honeysuckle belongs to a family of plants with no New Zealand native species within it;

 The two submissions from Māori provide tacit support for the application;

 Alternative control tactics are not explained or offered by the applicant; and

 The benefits to the Māori economy are not stated.

7.25. Ngā Kaihautū recommends that:

 The decision-making committee considers this application taking into account the issues outlined above and any mitigation measures as appropriate. 7.26. As mentioned in section 3.14 of this report Te Rūnanga o Ngāi Tahu submitted that they support this application because of their concern about the weed’s damaging effects on native species.

7.27. The Huakina Development Trust expressed concern about the possibility of the organism becoming a pest itself and were curious about the ease with which it could be eradicated if the need arose.

7.28. However, given that the organism is unlikely to cause damage to, or compete with native species, we are confident that it is unlikely to become a pest and require eradication.

Identification of potential adverse effects to kaitiakitanga

7.29. Māori perceptions of the natural world have been formed over hundreds of years where their existence has depended on their ability to maintain a symbiotic relationship within it. This reliance on each other was so deep-rooted that Māori not only consider the environment to be a taonga (treasure) but they also consider it to be their tupuna (ancestor) from where all things originate and descend. Therefore, Māori have the obligation as kaitiaki (guardians) to ensure their ancestor maintains a high level of wellbeing and that future generations are able to enjoy their relationship with papatuanuku in the same manner that contemporary Māori do.

7.30. Concerns have been raised over many years that biological control agents would upset the natural balance that exists in nature, so Māori have continuously sought assurances that the release of each biological control poses no threat to taonga species and the ecosystem.

7.31. In response, and as discussed in greater detail in the application, the applicant has stated that any adverse effect on native species is considered highly unlikely as the agent is host specific.

7.32. Given this assessment we anticipate a minimal effect on taonga species and iwi/Māori kaitiaki responsibilities to be highly improbable. The level of effect is therefore deemed to be minimal.

. June 2013 19

EPA advice APP201710

Impact on the principles of the Treaty of Waitangi (Te Tiriti o Waitangi)

7.33. Under section 8 of the Act, all persons exercising powers and functions under the Act are to take into account the principles of the Treaty of Waitangi (te Tiriti o Waitangi).

7.34. In reference to the “principles” of the Treaty of Waitangi, as currently accepted by the Courts and Waitangi Tribunal, they are stated to be that of partnership, participation and protection.

7.35. The principles of partnership and participation refer to the shared obligation on both the Crown and iwi/Maori to act reasonably, honourably and in good faith towards each other to ensure the making of informed decisions on matters affecting the interests of Māori. In fulfilment of these principles, as previously stated, the applicant has completed a consultation program for the application and responded to queries.

7.36. The principle of active protection refers to the Crown‘s obligation to take positive steps to ensure that Māori interests are protected. Taking into account this principle requires this application to provide sufficient evidence to show that the introduction of L. glorifica does not pose a significant risk to native or taonga species, ecosystems and traditional Māori values, practices, health and well-being.

7.37. We consider that the application provides sufficient information to take into account the principle of “active protection” and is considered to be consistent with the principles of the Treaty of Waitangi. Given this assessment we anticipate a minimal effect on the principles of the Treaty of Waitangi and therefore the level of effect is therefore deemed to be negligible.

Conclusion on adverse and positive effects

7.38. After completing our risk assessment and reviewing the relevant information, we consider that the adverse effects of releasing L. glorifica are negligible and the positive effects are non-negligible. Therefore the positive effects from the import and release of L. glorifica outweigh the adverse effects. We therefore recommend that the import and release of Limenitis glorifica be approved.

Asela Atapattu Thomas Haapu Dr Kate Bromfield Manager Senior Advisor Senior Advisor New Organsims Māori and Policy New Organisms

. June 2013 20

EPA advice APP201710

References

Barron M.C. Barlow, N.D. and Wratten, S.D. (2003). Non-target parasitism of the endemic New Zealand red admiral butterfly (Bassaris gonerilla) by the introduced biological control agent Pteromalus puparum. Biological Control 27: 329-335

Briese D. (2005). Translating host-specificity test results into the real world: The need to harmonize the yin and yang of current testing procedures. Biological Control 35: 208–214.

Marohasy J (1998). The design and interpretation of host-specificity tests for weed biological control with particular reference to insect behaviour. Biocontrol News and Information 19: 13N–20N.

Paynter Q., Fowler, S.V., Gourlay, A.H., Groenteman, R., Peterson, P.G., Smith. L. and Winks, C.J. (2010). Predicting parasitoid accumulation on biological control agents of weeds. Journal of Applied Ecology 47: 575–582

Tanaka, B. (1978). Larval food-plants and distribution of Japanese Lagoda (Lepidoptera: Nymphalidae). Transcripts of the Lepidopteron Society of Japan. 29(1): 35-45.

Wapshere A.J. (1974). A strategy for evaluating the safety or organisms for biological weed control. Annals of Applied Biology 77: 201–211.

Williams, P.A., and Timmins, S.M. (1998). Biology and ecology of Japanese honeysuckle (Lonicera japonica) and its impacts in New Zealand. Science for Conservation 99. Department of Conservation.

Williams, P.A.; Timmins, S.M.; Smith, J.M.B.; Downey, P.O. (2001). The biology of Australian weeds 38 Lonicera japonica (Thunb) Plant Protection Quarterly 16(3): 90-100.

. June 2013 21

EPA advice APP201710

Appendix 1

Table 1 Summary of submissions received through public and targeted notification

Submitter/ Support/ Submission Submitter comments organisation Oppose

102832 Hawkes Bay Regional Support  L. japonica is a vigorous climbing vine that smothers Council desirable plants and shrubs  Can no longer be controlled by conventional means

102833 Donovan Scientific Neither  Concern over the impact on bumblebees Insect Research support nor oppose  Lonicera may be an important foraging plant

102834 Kiwifruit Vine Health Neither  Concerns over the adequacy of host testing on non- support nor native but valued plants, particularly those of value to oppose industry

102835 Northland Regional Support  L. japonica has the potential to disperse from its Council current locations

102837 Huakina Development Oppose  Support any move to reduce chemical use Trust  Threats to natives not adequately covered in application  Propose that the applicant discusses eradication if Limenitis becomes a pest

102840 Te Rūnanga o Ngāi Support  Risks to other flora and fauna low Tahu  Potential beneficial impact on the weed and reduction of the weeds damaging effects on native species

102848 Rob Morton Support  Japanese honeysuckle gets tangled with native vegetation  Hand clearance is difficult  Biocontrol would boost native bush restoration efforts

102882 Bay of Plenty Regional Support  Japanese honeysuckle is a widespread invasive pest Council in the Bay of Plenty Region  Successful biological control will reduce the need and cost associated with control of Japanese honeysuckle

102900 Cliff Mason Oppose  The introduction of alien species damages the ecology of New Zealand  Inadequate understanding of the impact of L. japonica and its removal on the communities that it invades  No relevant information on the effects of L glorifica herbivory on L. japonica  Host testing is inadequate, especially as it includes no native species  The theoretical basis of the applicant's methods and . June 2013 22

EPA advice APP201710

Submitter/ Support/ Submission Submitter comments organisation Oppose

conclusions are artefactual and unstable  The proposal is an experiment conducted upon New Zealand’s environment  The principal vector in the dispersal of L japonica is human activity and changing this is preferable to changing the ecology of New Zealand  Many of the places in which L. japonica is a pest have a multitude of pest plant species and control of one is unlikely to be of ecological benefit  Does not wish the experience of nature to involve encounters with introduced species

102902 Waikato Regional Support  Japanese honeysuckle is a serious weed and a true Council intractable species in the Waikato.  It is widespread and has not reached its true potential.  Managing this plant is difficult and requires vigilance  Biocontrol may help reduce the use of herbicides, and reduce the invasiveness of Japanese honeysuckle

102904 Thames Coromandel Support  Japanese honeysuckle is a pest plant in the Thames District Council Coromandel District  Support for the proposal is given on the proviso that testing is thorough and no release of the biological control agent is undertaken if harm to native or beneficial introduced species could occur

102920 Moths and Butterflies of Neither  Concern about the effects on nature New Zealand Trust support nor  Doubt about biocontrol efficacy oppose  Appreciation of the way butterflies enhance natural experiences and awaken wonder in children

. June 2013 23

EPA advice APP201710

Appendix 2

DOC comments on EPA new organism for release application

8th May, 2013

Application number: APP201710 Applicant: Greater Wellington Regional Council Application purpose: To import and release the Honshu white admiral butterfly (Limenitis glorifica) for the control of Japanese honeysuckle (Lonicera japonica) Submission closes: 18 June 2013

Thank you for the opportunity to comment on this application. We do not wish to be heard at a public hearing in support of our comments.

DOC and the National Biological Control Collective The Department of Conservation is a member of the National Biological Control Collective, along with the regional councils (and Landcare Research as the science provider/adviser). The Department provides funding to the Collective to assist with the biological control of weeds programme. The Department of Conservation supports this application to import and release the Honshu white admiral butterfly (Limenitis glorifica) for the control of Japanese honeysuckle (Lonicera japonica).

Assessment of risk to conservation values The Department has no reason to dispute the results of the host range testing and it is our opinion that the Honshu white admiral butterfly (Limenitis glorifica) poses a negligible risk to the native flora and fauna of New Zealand.

Japanese honeysuckle (Lonicera japonica) is a significant environmental weed and contributions to successful control via a biocontrol agent will help to protect conservation values. Please refer to our comments provided directly to the applicant during earlier consultation (docdm-1151392) for further detail.

Comments co-ordinated on behalf of the Department of Conservation by: Verity Forbes Technical Advisor Biosecurity Threats Science and Technical Group, Wellington

Contributors: Tom Belton, Technical Advisor Threats, Science & Technical, Hokitika Chris Green, Technical Advisor Threats, Science & Technical, Auckland

. June 2013 24

EPA advice APP201710

Appendix 3

Response to the review by Dr T. Withers of the application to introduce Honshu white admiral

Quentin Paynter and Richard Hill - 12 April 2013

1. ‘Did Landcare Research consider whether some other closely related native plants should be tested as an out-group, even if they do not share the form of a “honeysuckle” vine?

We did consider including native plant species during the development of the test plant list. However, the inclusion of native species was considered unnecessary because Japanese honeysuckle is so taxonomically remote from the most closely-related native species. It is clear that NZ natives are not at risk as they are all very distantly-related (different plant Orders) to Japanese honeysuckle. If ornamentals had not been an issue, a case could have been made that enough was known about the host-range of L. glorifica to import it without additional host-range testing. Our testing rules out non-target attack in all families/sub-families of the Order except to members of the Caprifoliaceae sub-families Diervilloideae and Caprifolioideae and shows risk is low for key ornamentals (L. nitida, Weigela).

2. ‘The first series of tests utilised individual larvae, each as an independent replicate each on its own severed leaf sitting on a moistened piece of filter paper. So n = 20 for each of the host plant tests run for 10 days. I see no reason why this first set of data could not be analysed statistically.

‘As a result of this method the data on larval survival on each plant is one replicate and so statistical analysis on the resultant larval survival to pupa or adult cannot be undertaken. It makes it more difficult to draw conclusions from this insufficiently replicated and not statistically analysable data set’.

It was not considered necessary to present statistical analysis of the data in the application because the test results were clear cut, and the interpretation was self-evident. Nevertheless, in response to Dr Withers comments Dr Paynter has completed an analysis which is presented here. It should be stressed that this analysis does not change any aspect of the interpretation of the data presented in the application.

No choice L1 starvation and survival tests Analysis was done using the R statistical program (R Development Core Team 2008).

A Kruskal Wallis test revealed a significant effect of plant species on larval survival Kruskal-Wallis chi- squared =103.7899, p < 0.001). A post-hoc test using Mann-Whitney tests with Bonferroni correction showed significant differences between Japanese honeysuckle and Weigela and Japanese honeysuckle and Leycesteria (p < 0.05); and between Japanese honeysuckle and Abelia, Dipsacus, Morina, Valeriana & Viburnum (all p < 0.001). There was, however, no significant difference between Japanese honeysuckle and Symphoricarpus (P = 0.203).

Abelia Dipsacus Leycesteria Morina Symphoricarpus Valeriana Viburnum Weigela sp. sylverstri formosa longifolia sp. officinalis sp. florida

Lonicera *** *** * *** n.s. *** *** * japonica

. June 2013 25

EPA advice APP201710

20 a 18 16 14 12 ab 10 bc* 8 bc* 6 4

Number alive afterdays 10 2 c*** c*** c*** c*** c*** c***

0

Abelia sp.

Viburnum sp.

Lonicera nitida

Weigela florida

Morina longifolia

Lonicera japonica

Dipsacus sylvestris

Valeriana officinalis

Symphoricarpus sp. Leycesteria formosa Adoxaceae Diervilloideae CaprifolioideaeCaprifolioideaeCaprifolioideaeCaprifolioideae Linnaeoideae Morinoideae Dipsacoideae Valerianoidea

Whole plant tests Counter to Dr Withers’ comment this test was replicated, as there were two replicates per plant species (each with 5 larvae), and not one replicate with 10 larvae as Dr Withers assumed. Dr Paynter considered the data was capable of statistical analysis. A Bartlett’s test indicated homogeneity of variances in the numbers of larvae surviving (out of five) across groups (test plants), so an Analysis of Variance was performed to investigate whether the number of larvae that completed development (out of 5) per replicate varied between test plant species. With only 2 replicates per plant species (5 species), replication was indeed low, but there was, nevertheless a significant effect of plant species on the number of larvae developing to adult (P < 0.05).

Accumulated analysis of variance Change d.f. s.s. m.s. v.r. F pr. + rep 1 0.000 0.000 0.00 1.000 + Plant_species 4 27.600 6.900 6.90 0.044 Residual 4 4.000 1.000 Total 9 31.600 3.511

Post-hoc analysis (Fischer’s Least Significant Difference) found survival on Japanese honeysuckle and Leycesteria was not significantly different, but survival on Japanese honeysuckle was higher versus the other test plants (see graph).

. June 2013 26

EPA advice APP201710

6

a 5

4 a

3

2 Number surviving to adult (out of(outNumber5)adult surviving to b b b 1

0 Weigela florida Lonicera japonica Symphoricarpus Heptacodium Leycesteria sp. micanoides formosa Test plant species

Dr Wither’s expressed some concern that additional larvae were added to ensure 10 larvae were used per plant species in the whole plant tests. She interpreted this to mean that some larvae had changed host plant during their development. This is not the case – the additional larvae were all newly emerged L1 larvae that had not fed on other plant species (so previous experience of other host-plants will not have altered the test result). In any case, the numbers of added larvae were small (2 for Weigela and 3 for Leycesteria) and would not have had a major influence on the overall proportions (e.g. for Weigela, 1 larva survived to adult and 1/10 = 10%, while 1/8 = 12.5%).

References R Development Core Team (2010) R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org/.

. June 2013