NOR06005.Pdf(PDF, 783

Home , Apioninae, Cassida, Ceratapion onopordi

ER-AF-NOR-1-2 09/05

FORM NOR

Application for approval to

IMPORT FOR RELEASE OR RELEASE FROM CONTAINMENT ANY NEW ORGANISM INCLUDING A GENETICALLY MODIFIED ORGANISM BUT EXCLUDING CONDITIONAL RELEASE AND RAPID ASSESSMENT

[Short title is: New Organism Unconditional Release]

under section 34 of the Hazardous Substances and New Organisms Act 1996

Application Title: Import and release Ceratapion onopordi and Cassida rubiginosa, two biological control agents for Californian thistle

Applicant Organisation: Californian Thistle Action Group ERMA Office use only

Application Code: Formally received:____/____/____

ERMA NZ Contact: Initial Fee Paid: $

Application Status:

20 Customhouse Quay, Cnr Waring Taylor & Customhouse Quay PO Box 131, Wellington Phone: 04-916 2426 Fax: 04-914 0433 Email: [email protected] Website: www.ermanz.govt.nz

IMPORTANT

1. An associated User Guide is available for this form. You should read the User Guide before completing the form. If you need further guidance in completing this form please contact ERMA New Zealand.

2. This application form covers importation for release or release from containment of any new organism (i.e. full or unconditional release) including genetically modified organisms but excluding conditional release and rapid assessment, under section 34 of the HSNO Act 1996.

If you are making an application to import for release or release from containment any new organism with controls (i.e. conditional release) use Form NOCR. If you are making an application to import for release a new organism that is not a genetically modified organism by rapid assessment use Form NO1R. If you are making an application to field test a genetically modified organism use Form NO-04.

3. The form replaces all previous versions of Form NOR. Old versions should not now be used. You should always check with ERMA New Zealand or on the ERMA New Zealand website for the most up-to-date versions of this form.

4. You can talk to an Applications Advisor at ERMA New Zealand who can help you scope and prepare your application. We need all relevant information early on in the application process. Quality information up front will speed up the process and help reduce costs.

5. This application form may be used to seek approvals for importing or releasing more than one new organism where the organisms are of a similar nature. 6. Any extra material that does not fit in the application form must be clearly labelled, cross-referenced, and included as appendices to the application form.

7. Commercially sensitive information must be collated in a separate Appendix. You need to justify why you consider the material commercially sensitive, and make sure it is clearly labelled as such.

8. Applicants must sign the form and enclose the correct application fee (plus GST). The initial application fee can be found in our published Schedule of Fees and Charges. Please check with ERMA New Zealand staff or the ERMA New Zealand website for the latest schedule of fees. We are unable to process applications that do not contain the correct initial application fee.

9. Unless otherwise indicated, all sections of this form must be completed for the application to be progressed.

10. Please provide an electronic version of the completed application form, as well as sending a signed hard copy. Until a signed hard copy of the application is received, ERMA New Zealand will not be able to process your application.

11. Note: Applications for full (unconditional) releases (this form) shall be publicly notified by the Authority under section 53(1)(b) and may go to a hearing pursuant to section 60 of the Act. You can get more information by contacting us. One of our staff members will be able to help you. This application form was approved by the Chief Executive of ERMA New Zealand on 20 September 2005. ERMA New Zealand 20 Customhouse Quay, PO Box 131 Wellington, NEW ZEALAND Telephone: 64-4-916 2426 Facsimile: 64-4-914-0433 E-mail: [email protected] www.ermanz.govt.nz

Section One – Applicant Details

1.1 Name and postal address in New Zealand of the organisation or individual making the application:

Name Californian Thistle Action Group

Postal Address c/o Clutha Agricultural Development Board PO Box 149 BALCLUTHA

Physical Address Development house 6 John St BALCLUTHA

Phone (03) 418 3188

Fax (03) 418 4201

E-mail [email protected]

1.2 If application is made by an organisation, provide name and contact details of a key contact person at that organisation This person should have sufficient knowledge to respond to queries and have the authority to make decisions that relate to processing of the application.

Name Malcolm Deverson

Position Secretary, Californian Thistle Action Group

Address c/o Clutha Agricultural Development Board PO Box 149 BALCLUTHA

Phone (03) 418 3188

Fax (03) 418 4201

E-mail [email protected]

Alternative contact Name Richard Hill

Position application author

Address Richard Hill & Associates Ltd Private Bag 4704, CHRISTCURCH

Phone (03) 325 6400

Fax (03) 325 2407

E-mail [email protected]

1.3 If the applicant is an organisation or individual situated overseas, provide name and contact details of the agent authorised to transact the applicant’s affairs in relation to the application This person should have sufficient knowledge to respond to queries and have the authority to make decisions that relate to processing of the application.

Name Not applicable

Position Not applicable

Address Not applicable

Phone Not applicable

Fax Not applicable

E-mail Not applicable

Section Two – Purpose of the Application and Reasons for Requesting a Full (Unconditional) Release This form is to be used for a standard (publicly notified) application (i.e. other than by rapid assessment), to import for release, or release from containment, any new organism (including a genetically modified organism). It is not intended to cover conditional releases.

2.1 Give a short summary statement of the purpose of this application to be used on ERMA New Zealand’s public register – Maximum 255 characters (including spaces and punctuation) Briefly describe the organism(s) to be imported for release or released from containment and the purpose(s) for which you wish to release the organism(s).

Note: An organism is „released‟ when it is not required to be held in a containment facility registered by the Ministry of Agriculture and Forestry. Once released it is no longer considered a new organism.

Approval is sought by the Californian Thistle Action Group to import and release two insects, Ceratapion onopordi (Brentidae) and Cassida rubiginosa (Chrysomelidae), for biological control of the weed Californian thistle (Cirsium arvense).

2.2 Provide a short description of the background and aims of the proposal suitable for lay readers Describe in less than one page the rationale for the proposal to release these organisms, including the potential use for the organism(s), so that people not directly connected with the research can understand the reasons for the release.

The Californian Thistle Action Group (CalTAG) wishes to introduce two biological control agents for Californian thistle. Californian thistle occurs throughout New Zealand and is one of the most serious and persistent weed problems of pastoral and cropping farms. Unlike other thistles, Californian thistle has deep, creeping roots that spread outward and generate dense stands of prickly stems. Stands out-compete other plants and deter animals from grazing. Managing the weed well is difficult, expensive, and often uneconomic. Two estimates put the annual cost imposed on land managers by Californian thistle in Otago and Southland alone at $32 and $67 million.

The CalTAG project seeks to implement a sustainable approach to the management of Californian thistle on pastoral farms initially in Otago, Southland, Manawatu and Wanganui. CalTAG comprises local farmers, and representatives of Environment Southland, Horizons Manawatu-Wanganui, Clutha Agricultural Development Board, FRST, The Foundation for Arable Research, Landcare Research, AgResearch, and Meat and Wool Innovation. The Group was formed in 1999.

Part of the CalTAG project involves the introduction of biological control agents that will tilt the playing field, and make existing management techniques more effective, efficient and safe. Two insects will be introduced. Ceratapion onopordi is a minute weevil from Europe whose larvae feed on and can debilitate the stems of a range of thistles. The larvae of Cassida rubiginosa feed on the leaves of a range of thistles, reducing the growth rate and survival of plants. This species was accidentally introduced to North America over 100 years ago. Studies there show that it does not attack valued plants, and that it can reduce the damage caused by Californian thistle.

The biocontrol agents have been extensively tested in the laboratory and in field tests in Europe to make sure that both are safe to release in New Zealand. Most insects selected as biocontrol agents are chosen because they only feed on the

Application for approval to import for release or release ER-AF-NOR-1-2 09/05 from containment any new organism including a genetically modified organism but excluding conditional Page 6 release and rapid assessment, under section 34 of the Hazardous Substances and New Organisms Act 1999 target plant. In this case neither species is restricted to a single host. Both feed on a narrow range of species belonging to the thistle tribe (Cardueae). Neither can feed on any plants outside that tribe. These species can be safely used in New Zealand because there are no native plants in the tribe Cardueae. Two potentially valuable crops, safflower and globe artichoke, belong to this tribe but tests and field records suggest that these species are not at significant risk. Laboratory tests indicate that several related ornamental species including cornflower and burdock could be attacked. However, Cassida is not considered a pest of these species in North America, and so attack is likely to be insignificant. The culturally important puha species belong to a related tribe of the daisy family. These species have been extensively tested, and are not at risk from either insect.

Four previous control agents have been introduced to attack Californian thistle, but none have been effective. As with any biological control project, whether these new species will provide adequate control of Californian thistle is uncertain. Nevertheless, field evidence and economic analysis of the proposal suggest that proceeding with the introduction of these species is justified.

2.3 Set out the reasons for this application being for a full (unconditional) release rather than for a conditional release Set out the reasons for this application being for full (unconditional) release rather than for conditional release. Under section 38B of the HSNO Act the Authority may consider an application for full (unconditional) release as if it were for conditional release (i.e. conditions can be set), with the agreement of the applicant. You should provide sufficient information to enable the Authority to decide whether or not it should approach you about obtaining agreement to switch from full (unconditional) to conditional release.

Application is sought for full (unconditional) release of both species.

Although the initial releases are to be made in Otago and Southland, it is intended and expected that these two agents will colonise all regions of New Zealand, either naturally or with the assistance of farmers and/or regional councils.

The applicants consider that sufficient information is presented with this application to allow the Authority to adequately assess the risk of introduction without requiring additional information that might come from conditional release.

Insects with a narrow host range may exhibit subtle variations in host range between populations, and it is sometimes important to source introduced agents from the population used in host-range tests. This is not an important issue in this case because the wider host range of the agents is already acknowledged and because there are no close relatives to the thistles among the related native flora. The agents will be sourced from the same regions as the insects used in recent host-range tests, but the applicants do not feel it is necessary to constrain the source of agents that can be introduced.

Section Three – Information on the Organism(s) to be Released and any Inseparable Organisms If the application is for release of more than one organism, information must be provided separately for each organism. If there are commercial reasons for not providing full information here alternative approaches must be discussed with and agreed by ERMA New Zealand.

3.1 State the taxonomic level at which the organism(s) to be released are to be specified If the taxonomic level is higher or lower than “species”, provide reasons for this. The reasons should take account of the need to adequately describe the risk.

Both species will be designated at the species level

3.2 Give the unequivocal identification of the organism(s) to be released Please provide details of the following:

Latin binomial, including full taxonomic authority (e.g. ----- Linnaeus 1753) class, order and family:

Species 1.

Class: Insecta Order: Coleoptera Family: Brentidae Subfamily: Apioninae Tribe: Ceratapiini Genus: Ceratapion Subgenus: Acanephodus Species: onopordi (W. Kirby, 1808)

Common name(s), if any: Thistle stem weevil Type of organism: Insect, weevil Strain(s) and genotypes(s), if relevant: From Europe Other information, (e.g. information on consideration of the organism(s) by other states, countries or organisations):

This is a valid species, but there is debate about the genus name. Opinion is coalescing around the use of Ceratapion but the species has also been known variously in recent years as Apion onopordi and Acanephodus onopordi. The subfamily Apioninae is sometimes placed in the family Apionidae (e.g. the Fauna Europaea, www.faunaeur.org), but New Zealand usage is employed here (Kuschel 2003, R. Leschen, Landcare Research, Appendix A). This species has not been intentionally released in any other country.

Species 2.

Class: Insecta Order: Coleoptera Family: Chrysomelidae Subfamily: Cassidinae

Tribe: Cassidini Genus: Cassida Subgenus: Cassida Species: rubiginosa O.F.Müller, 1776

Common name(s), if any: Green thistle beetle Type of organism: Insect, leaf beetle Strain(s) and genotypes(s), if relevant: From Europe Other information, (e.g. information on consideration of the organism(s) by other states, countries or organisations):

This species has not been intentionally released in any country for the purpose of biological control. However, it was accidentally introduced into North America over 100 years ago, and can now be found commonly in Northwest USA and neighbouring provinces of Canada. There it is known as a defoliator of a range of thistle species (Section 3.4).

3.3 Provide unique name(s) for the new organism(s) to be released These name(s) will be on the public register and should clearly identify the organism.

Thistle stem weevil, Ceratapion onopordi (Kirby), (Coleoptera, Brentidae) Green thistle beetle, Cassida rubiginosa Müller (Coleoptera, Chrysomelidae)

Specimens of both species will be lodged with the National Arthropod Collection, Landcare Research, Auckland, and with the MAF Reference Collection, Lincoln. Specimens will be submitted to Dr R. Leschen of Landcare Research to confirm identification before release.

3.4 Characteristics of the organism(s) to be released Provide information on the biology, ecology and the main features or essential characteristics of each organism(s) to be released. Provide information on affinities of the organism(s) with other organism(s) in New Zealand. You should also indicate whether the organism(s) is pathogenic or a potential pest or weed. This information should be relevant to the identification of the risks of the organism(s) (section 6 of this form).

Ceratapion onopordi

Description and biology Adult: The adult is a small (3–4 mm) dull-black weevil. Larva and pupa: Both stages occur within the stem. Larvae are minute.

Developmental characteristics and requirements The weevil overwinters as an adult, probably in the leaf litter. In spring females feed on thistle leaves, drill out a chamber in the stem at ground level, and lay a single egg in each. Larvae mine at ground level within the collar, and in the stem before pupating there. Little is known about the developmental rates of this insect.

Native distribution, habitat and climate requirements C. onopordi is present throughout Western Europe

(shaded = known to be present)

Establishment and dispersal This insect has not been used for biological control before, and is not known to be established outside of its native range.

Host records Gassman (2005) has reviewed all known European records. C. onopordi is not specific to one host. It typically feeds on Cirsium and Carduus species, but has not been recorded on any species outside the tribe Cardueae (thistles).

It is not known as a principal pest of the valued „thistle‟ species safflower and globe artichoke in Europe (http://www.inra.fr/Internet/Produits/HYPPZ/crops.htm), and is not recorded there as a garden pest.

Impact and ecology C. onopordi females appear to discriminate previously infested stems, and usually avoid laying more than one egg per stem. In most cases one larva will not kill a stem. However, infestation significantly reduces stem and root vigour, and suppresses the ability of the stem to compete with other vegetation (Friedli & Bacher 2001). C. onopordi promotes systemic infection by the European rust fungus Puccinia punctiformis (Str) Röhl (already present in New Zealand) in the year following weevil infestation, leading to death of stems. There is a clear synergy between weevil attack, plant competition, and later plant pathogen attack (Friedli & Bacher 2001). The issues surrounding this interaction have been summarised and assessed by ERMA (2004).

Affinities with New Zealand fauna Seven species belonging to the family Brentidae occur in New Zealand. Six of these belong to the subfamily Apioninae but none to the genus Ceratapion. The species of the tribe Apionini in New Zealand include the introduced gorse seed weevil Exapion ulicis (Forster, 1771), and the other apionines are found on the trees Nothofagus sp., Libocedrus bidwillii and Metrosideros sp. It is unlikely that C. onopordi will compete with these species here (R. Leschen, Landcare Research, pers. comm.; Appendix A) or occur n the same habitats.

Two Trichomalus spp. are the only parasitoids known to attack Ceratapion onopordi in Europe (e.g.Freese 1995). The genus Trichomalus is not known to be present in New Zealand (J. Berry, Landcare Research, pers. comm.; Appendix A). The introduced parasitoid Microctonus aethiopoides will potentially invade grassland ecosystems. This attacks a small proportion of gorse seed weevil. This is a weevil of similar size to Ceratapion onopordi but is unlikely to have a major impact on populations of the new agent (B. Barratt, AgResearch, pers. comm.; Appendix A).

Cassida rubiginosa

Image from www.cirrusimage.com/

Much of the information presented here is from Harris (2005).

Description and biology Adult: The adult is an oval green beetle, 6.0 to 7.5 mm long, with a hard covering on the head, thorax and abdomen that forms a carapace extending beyond the body like the shell of a tortoise. There is a brown spot at the centre top of the wing covers and the underside is black. Females are slightly larger than males.

Larva: The larva has prominent lateral spines and a forked tail spine on which it accumulates moult skins and excrement. This is held as a protective parasol over the insect's back, although some parasite species orient to the parasol.

Pupa: The pupa is a flat brown oval with a margin of black ray-like spines. It is attached to a leaf or stem by its tail segments and if disturbed it flaps like a hinge.

Developmental characteristics and requirements The eggs are laid in cycles of approximately 6 weeks with 7-week rests. Potentially, a total of 1000 can be laid on the undersides of leaves, usually in groups averaging 4.6 that are normally enclosed in a case (ootheca). The eggs can survive frost, but do not develop below 10°C. At 27°C they hatch in 6 days. There are five larval instars with the first three window-feeding, principally on the underside of leaves, and the last two making holes from the upper side of the leaf. At 27°C the pupal stage lasts about 4 days.

There is one generation a year in Canada.

The adults overwinter in soil litter. These emerge on warm days at the end of winter. They feed on the first thistle leaves to appear in the spring, making round holes similar to those of the larvae. Feeding on nodding thistle starts about a month earlier than on Californian thistle because the foliage is available earlier. The young adults feed on thistle leaves, but around seed dissemination they leave for the tender foliage of young rosettes. Here they continue feeding until the foliage hardens shortly before winter. The adult lifespan is about 80 weeks with an egg-to-adult developmental period of 6 weeks.

Native distribution, habitat and climate requirements This beetle occurs throughout the Palearctic (all of Europe, eastern Mediterranean, North Asia, North Africa).

Establishment and dispersal This species was accidentally introduced to Canada and causes local defoliation of thistles in the provinces of Nova Scotia, Prince Edward Island, Quebec, and Ontario. It was first found in Quebec in 1901. In Virginia the beetle was established with releases of 150–200 adults and dispersed 30–150 m a year.

Cassida rubiginosa has not been redistributed to other parts of North America because it attacks North American native thistles.

Host records Gassman (2005) has reviewed all known European records. C. rubiginosa is not specific to one host. It feeds on a range of thistles, but has not been recorded on any species outside the tribe Cardueae (thistles). In Ontario the beetle attacks Cirsium arvense , C. vulgare, Carduus nutans, C. acanthoides , and occasionally Arctium minus and Centaurea jaceae. In Virginia it also occurs on the native thistle Cirsium discolour (Harris 2005). It is not a principal pest of the valued „thistle‟ species safflower and globe artichoke in Europe (http://www.inra.fr/Internet/Produits/HYPPZ/crops.htm).

In field studies in the United States, Kok et al. (2000) showed that the beetle laid eggs equally freely on nodding and Californian thistle. Development of eggs to adult stage was completed in 27.4 days on Californian thistle, 1.1 day shorter than that on musk thistle. However, survival of larvae when feeding on nodding thistle (Carduus nutans) was only half of that on Californian thistle (Kok et al. 2000).

Impact and ecology Ang et al. (1995) found that in dry years in Virginia, five parasite-free beetles per plant reduced aboveground dry weight of Californian thistle by 88% in plots with grass and only 25% of the plants survived to the end of the second year. Biomass reduction in plots with only thistle was 62%. The impact was less in moist years, but still substantial.

Three parasitoids were accidentally introduced with the beetle, and these normally keep its density in Ontario below that needed to defoliate Californian thistle. Parasitism in southern Virginia is lower, normally 15–25%, and the thistle is reduced, particularly in sites with high grass competition. In a conference presentation summarising the known information about the biology of Cassida rubiginosa, Kok et al. (2000) reported that thistle growth was inhibited when defoliation exceeded 50% of the leaf tissue. When density exceeded 20 beetles per rosette, less than one-third of the plants survived. Studies on the competitive growth of Californian thistle, tall fescue, and C. rubiginosa indicated that the beetle and induced plant competition could be effective as part of a comprehensive approach for area-wide Californian thistle control.

Affinities with New Zealand fauna There are a large number of chrysomelid species in New Zealand, but no species of the tribe Cassidinae occur naturally here (R. Leschen, Landcare Research, pers. comm.; Appendix B).

There are 19 parasitoid species known from Cassida rubiginosa worldwide. The only one known to be present in New Zealand is the eupelmid wasp Eupelmus vesicularis (Retzius) (J. Berry, Landcare Research, pers. comm.; Appendix A). E. vesicularis accounted for 0.1% of the parasitism of Cassida rubiginosa in northern Virginia (Ward & Pienkowski 1978, Tipping 1993 in McClay 2002), but was not found by Ang et al. (1995) in south-western Virginia. E. vesicularis is a primary and secondary parasitoid with a wide host range (c.135 host species known), mostly in the families Curculionidae (primary) and Eurytomidae (secondary). It has been recorded from dipteran and lepidopteran hosts as well as from other coleopterans and hymenopterans. If C. rubiginosa is introduced as a control agent and establishes in New Zealand, it is possible that it will be attacked by E. vesicularis where the two species are sympatric. In New Zealand E. vesicularis is reasonably widely distributed over the South Island (J. Berry, Landcare Research, pers. comm.; Appendix A). The low parasitism rate of C. rubiginosa in northern Virginia may indicate that C. rubiginosa is not a preferred host, at least where there is an established parasitoid complex

The following parasitoid genera are present in New Zealand and contain species known to attack C. rubiginosa in other parts of the world: Anaphes, Aprostocetus, Brachymeria, Oomyzus, Pediobius and Tetrastichus. Related native species in these genera may be able to attack C. rubiginosa in New Zealand, but there is no way of knowing this. It is possible that other exotic species known to attack Cassida rubiginosa are present in New Zealand but have not yet been recorded, since the majority of the parasitic New Zealand hymenopteran fauna is undescribed or undetermined (J. Berry, Landcare Research, pers. comm.; Appendix A).

Interactions between proposed agents with other species introduced to attack Californian thistle in New Zealand Other biocontrol agents have been introduced to attack Californian thistles, but either did not establish, remain rare or are ineffective. For this reason, and because insects feed on different parts of the plant, no significant competition with other Californian thistle control agents is expected.

Damaging Type of Agent Plant part Season Status Impact stage attack

Altica carduorum Larva/adult Defoliation leaf Spring Not established Nil Lema cyanella Larva/adult Defoliation leaf Spring Rare Nil Not Hadroplonthus litura Larva Stem mine Root/stem ?? Nil established? Urophora cardui Larva Stem gall Stem Summer Established Uncertain Ceratapion onopordi Larva Stem mine Root collar/Stem Spring Not released - Cassida rubiginosa Larva/adult Defoliation Leaf Spring Not released -

3.5 Identify and characterise any inseparable organisms Inseparable organisms are those which are inherently associated with each main organism e.g. gut bacteria in an animal.

No inseparable organisms are known for either Ceratapion onopordi or Cassida rubiginosa.

Associated organisms Thirty mid-size immature stages of each species will be randomly selected from the rearing colony, smeared and submitted to Biodiscovery NZ (Dr P. Wigley) for examination for unwanted associated micro-organisms.

No imported life stages will be released. This will eliminate any risk that associated parasitoids or predators (H. Gourlay, Landcare Research, pers. comm.) might be released. Both species will probably be reared through a full generation in quarantine.

Puccinia punctiformis is a rust fungus with a worldwide distribution. It is common in Europe. It is essentially host- specific to Cirsium arvense but has been recorded on other thistles on rare occasions. Of the 25 New Zealand records, distributed from Auckland to Southland, in the Landcare Research mycology database, 23 are from this host plant nad two are from other thistles (http://nzfungi.landcareresearch.co.nz/html/data.asp?ID=&NAMEPKey=9385). Ceratapion onopordi is capable of transmitting the rust between thistle stems (Friedli and Bacher 2001), and it is possible that imported adult weevils could carry spores of this rust (see Section 7 for more information). ERMA NZ (2004) considered the risk of escape of any contaminating fungus during containment was very unlikely because weevils were contained, and because air was filtered. Further, the report also concluded that any accidental liberation of the fungus was very unlikely to have any significant effects on the environment as the species is already present here.

Standard quarantine practice will be employed to minimise the risk of releasing the rust once permission to release weevils is granted. Imported weevils will not be released. The weevils will be reared on healthy, lab-raised plants for one generation. Any host plants showing evidence of ill health or rust infection during that period will be destroyed. Weevils of the F1 generation will be held for 2 days without plant material before removal from quarantine. The risk of transmitting the disease across generations under this regime and of releasing viable spores or mycelium of a European genotype of P. punctiformis (or any other associated plant disease) from quarantine is considered very low.

3.6 If the organism to be released is a genetically modified organism, provide details on the development of the organism If the organism to be released is a genetically modified organism, state whether the development of the organism was carried out under a HSNO approval. If this was the case, provide the approval number and translate the relevant details to the headings below. If the genetically modified organism is to be imported for release, also provide this information on its development to the extent possible under the following headings:

The Ceratapion onopordi and Cassida rubiginosa to be introduced have not been genetically modified.

Identify the category of the host organism (i.e. category 1 or 2) and genetic modification (i.e. category A or B) involved in the development of the organism with reference to the HSNO (Low-Risk Genetic Modification) Regulations 2003. Please explain your characterisation. Not applicable

Vector system(s) used in development of the genetically modified organisms. Not applicable

Type and source of additional genetic material. Not applicable

Use of special genetic material: please complete this table by marking the correct box Not applicable

Yes No Were(was) native flora or fauna used as host organism(s)? Was genetic material from native or valued introduced flora and fauna used? If native flora and fauna were involved, were the species concerned endemic to New Zealand? Was human DNA or cell lines used that are of known Māori origin? Was genetic material obtained directly from human beings? If Yes, provide additional details below.

If the genetic modification involves DNA of human origin, provide details of where the material was obtained (including provenance and/or informed consent), and whether approval was obtained from an Ethics Committee, and/or whether consultation with Māori has taken place. Consultation with Māori will only be required if the human DNA or cell lines are of known Māori origin. Where consultation has occurred, ensure you append any relevant information to the application including consultation feedback, minutes of meetings or other correspondence.

Not applicable

Other relevant details (such as what techniques or experimental procedures were used, whether any unusual manipulations were carried out, and how the foreign genetic material is expressed).

Not applicable

3.7 Does the organism have any other HSNO containment approvals not covered in 3.6 (e.g. import into containment, field test, or conditional release approval)? State whether the organism(s) to be released has any other HSNO containment approvals. If this is the case, provide the approval number(s) and give brief details of those approvals.

An application (NOC02004) to import Ceratapion onopordi (as Apion onopordi) into containment for further evaluation was approved by ERMA, approval code NOC002279. Import Permit no. 2006028253 was issued by MAF and a population is presently held in the Landcare Research containment facility at Lincoln.

Cassida rubiginosa has not been the subject of any previous applications.

Section Four – Establishment and Eradication of a Self-Sustaining Population Information under this and the next heading is required so that the Authority can take account the matters set out in section 37 of the Act.

4.1 Ability of organism(s) to establish a self-sustaining population Describe any ability of the organism to establish a self-sustaining population. This may include, but not be restricted to, information on the time taken for the organism(s) to become established, the likely geographical spread of the organism(s), and effects of variations in climate and altitude on the establishment, distribution, abundance and biology of the organism(s). Explain any situations where the establishment of this(these) organism(s) as a self-sustaining population would not be undesirable, along with the likelihood of each possibility. For a full (unconditional) release the issue of (un)desirability is crucial, because in many cases the establishment of a self-sustaining population will be expected (e.g. a bio-control release).

The objective of introducing Cassida rubiginosa and Ceratapion onopordi to New Zealand is to establish self- sustaining populations on thistle infestations, initially in the south of the South Island. It is expected that the insects will eventually colonise thistle populations throughout New Zealand, contributing to the suppression of thistle populations and the maintenance of weed control everywhere. It is not expected that any populations will be considered undesirable because both species pose insignificant risk to native plant populations, to the current level of thistle control, and to the integrity of native ecosystems and processes (see section 7).

4.2 Ease of eradication of a self-sustaining population Information under this heading must be provided although ERMA New Zealand understands this application seeks approval to import an organism for release or release from containment. Describe the ease of eradiation of the population, including the methods to be used, likely costs, and the likelihood of total eradiation of the population.

Not applicable

Section Five – The Proposed Release Programme (and Monitoring) Provide full details of your intended release programme e.g. information on the breeding and culture, and the life-stage and number of the organisms to be released; timing and location(s) of release etc. Also provide information on any post-release monitoring you intend to carry out, and why.

Rearing and release The populations of Ceratapion onopordi and Cassida rubiginosa to be released in New Zealand will be collected at or near the location from which insects were drawn for field host-range tests conducted in Europe (Gassman et al. 2005). Before release from containment, the identity of the populations to be released will be confirmed by Coleopterist Dr Rich Leschen of Landcare Research.

Cassida rubiginosa and Ceratapion onopordi will be shipped to the containment facility at Landcare Research, Lincoln, New Zealand. Once post-importation requirements have been met, and permission to release from containment has been granted by MAF, the insects will be removed from the quarantine facility for mass-rearing by Landcare Research staff on behalf of the applicant. Both insects will be reared on potted plants in cages at Landcare Research, Lincoln. Adults of succeeding generations will be collected from those plants in spring and released onto thistle-infested pastures. Initial releases of at least 100 weevils per release will be made in Southland and Otago. The size and number of releases of each species and the timetabling of releases will depend on how successfully each can be mass-reared.

Post-release monitoring Current funding for this project does not extend beyond June 2008. Landcare Research has undertaken to monitor the establishment of the two species until March 2008. One year after release, each site will be visited and all thistle plants within a 20-m radius of the release point will be examined for Cassida rubiginosa adults or characteristic leaf damage. Twenty randomly selected thistles within that area will be sampled by beating the foliage over a white sheet, or by suction sampling, to detect adult Ceratapion onopordi.

All other herbaceous asteraceous plants within a 20-m radius will also be examined and sampled in the same way to check for damage to non-target plants.

Once established, it is likely that populations of both species will grow for some years before approaching maximum density. Until this happens, more detailed monitoring and evaluation cannot be justified. Significant investment to evaluate the field effectiveness of either agent before the agents achieved maximum potential would be inefficient and counterproductive. Any further steps to gain funding to monitor the performance of either species will be considered only once full establishment is confirmed.

Section Six - Identification of adverse and beneficial effects (risks, costs, and benefits)

This section must include information on the adverse and beneficial effects referred to in the HSNO Act. Adverse effects include risks and costs, and beneficial effects are described as benefits. For convenience adverse effects are at times referred to simply as risks. All effects should be described in terms of the magnitude of the effect if it should occur and the likelihood of occurrence. Monetary and non-monetary effects should be considered, and a comment should be included on the distribution of the adverse and beneficial effects across affected parties.

In this part of the form you are required to identify the potential adverse and beneficial effects (risks, costs and benefits) of the organism(s) in the context of the application.

A very broad approach should be taken to this, so that a wide range of possibilities is canvassed. In the first instance you are required to identify all potential adverse and beneficial effects (risks, costs and benefits) whether you consider them to be non-negligible or not. This should be carried out for inseparable organisms as well as for the principal organism. To do this effectively you should consider both the source of the risk (or hazard) and what is at risk (or area of impact). You should also consider the route (or exposure pathway) between the source and the area of impact.

Essentially what you should end up with is a very brief description of the potential adverse and beneficial effects (e.g. the potential for the pathogenic micro-organism (hazard) to have adverse effects on human health (area of impact) from consumption of the organism (exposure pathway). A more detailed assessment of these and other matters will be required in the next section (section 7).

Once you have considered all possibilities then you should clearly identify those potential adverse and beneficial effects (risks, costs and benefits) that are considered to be potentially significant and warrant further more detailed assessment (in section 7). If you consider that the effects identified do not warrant detailed assessment, explain why.

You can refer to the ERMA New Zealand Technical Guides “Identifying Risks for Applications” and “Risks, Costs and Benefits for Applications” for further information and guidance on completing this section. These are available from the ERMA New Zealand website or in hard copy on request. Please undertake your identification of risks, costs and benefits under each of the following headings (areas of impact) which reflect those matters referred to in Part II of the HSNO Act:

The following risks and benefits of the proposed introduction of Cassida rubiginosa and Ceratapion onopordi have been identified by: Pre-application consultation with Māori Pre-application consultation with ERMA NZ Pre-application consultation with the Department of Conservation Pre-application consultation with other stakeholders Brainstorming with five Landcare Research weed ecologists Including potential risks and benefits identified in previous applications.

The risks and benefits have been tabulated below. Only those risks and benefits that have been scored C or above using ERMA‟s risk assessment matrix are systematically addressed and assessed in Section 7. Those scoring A and B may not be mentioned.

The risks and benefits pertain to: 1. The introduction of populations of Ceratapion onopordi and Cassida rubiginosa into the New Zealand environment. 2. Establishment of successful biological control of Californian thistle leading to lower plant population density and vigour. 3. Damage to other thistle species.

6.1 Identification of potential effects on the environment (in particular on ecosystems and their constituent parts) Taking particular account of sections 5(a), 6(a) and 6(b) of the Act, list the environmental risks, costs and benefits associated with the organism(s) to be released, and any inseparable organisms. Risks, costs and benefits in this category include those relating to the life-supporting capacity of air, water, soil and ecosystems; the sustainability of native and valued introduced flora and fauna; the maintenance of natural habitats; the intrinsic value of ecosystems; New Zealand‟s inherent genetic diversity; and animal or plant health.

List potential adverse and beneficial effects, separately. If you decide that some of them are not potentially significant and do not require further assessment, indicate why you have reached this conclusion.

Risks, costs, and benefits:

Source of potential risk/benefit Likelihood Severity or Risk Comments advantage score

Environment

Adverse effects (risks and costs) Parasitoid, predator and disease Unlikely Minor D Effect not predicted, patchy, local, relationships within food webs adversely reversible, not predictable, see Section affected 7.1 Non-target feeding reduces native plant Highly Moderate C See Section 7.1 populations improbable Selecting populations other than those Improbable Minor B Variation in host range between tested leads to unpredicted non-target populations not expected as these are effects oligophagous insects. See Section 7.1 Significantly fewer native insects/bees Improbable Minor B No native bees rely on thistle flowers because of fewer floral sources (B. Donovan, beekeeper, pers. comm). There are other late spring floral sources in the agricultural landscape for butterflies and other species Populations of native herbivores reduced Improbable Minor B Thistles not common in native habitats. by predation or competition Native herbivores not common on thistles Candidates not predaceous. Environment damaged because wrong Highly Minor B Agent identity will be checked before agents released improbable and after importation to containment. See Section 5 Thistles replaced by worse weed Very unlikely Minimal B No replacement plant will occupy more ground; the most likely replacements are desirable plants

Fewer thistles leads to increased soil Improbable Minimal B Thistles not valued for soil conservation erosion Increased predation on native insects by Very unlikely Minimal B Agents will make up only a small wasp populations increased by eating proportion of available prey for wasps; Cassida larvae no significant increase in wasp populations expected Reduced invertebrate biodiversity in Improbable Minimal B Thistles will be replaced by other paddocks vegetation with its own fauna

Benefits Parasitoid, predator and disease Unlikely Minor D Effect not expected, patchy, local, relationships within food webs reversible, not predictable, see Section beneficially affected 7.1 Significant reduction in the environmental Improbable Minor B Like CT, other thistles do not have impacts of other thistles significant environmental effects. Gains are likely, but the beneficial effects will be small. See Section 7.1 More successful habitat restoration Improbable Minor B Californian thistle limits the success of restoration efforts on Quail Island, but this is unusual, thistles not usually important to natural processes Pest bird population reduced because less Improbable Minimal B Reduced food unlikely to significantly thistle seed as food affect bird populations or damage caused by birds Invertebrate biodiversity in tussock Improbable Minimal B Not much native habitat is thistle pastures improved infested, any effects patchy and local More native plants as thistles reduce in Improbable Minimal B Ditto native habitats Replacement plants more Improbable Minimal B Ditto environmentally friendly Less herbicide in soil, air and waterways Unlikely Minimal B True, but limited benefit as herbicides are applied briefly and rarely, and (ideally) using procedures that limit adverse effects to soil, water and air Reduced bee populations reduce weed Unlikely Minimal B Bee populations unlikely to be reduced? seed production and hence populations Few weed populations pollination- and damage limited? Seed production not linked to population density? See Section 7.1

6.2 Identification of potential effects on human health and safety (including occupational exposure) Taking particular account of section 6(c) of the Act, list any potential risks, costs and benefits to human health that may be related to the release of the organism(s) in New Zealand. Consider the impact on people associated with the release programme as well as the wider community.

List potential adverse and beneficial effects separately. If you decide that some of them are not potentially significant and do not require further assessment, indicate why you have reached this conclusion.

Risks, costs and benefits:

Source of potential risk/benefit Likelihood Severity Risk Comments or score advantage Public health Adverse effects (risks and costs) Allergenic responses to insects Highly Minimal A No such response known for beetles outside improbable houses Public fearful of insects Improbable Minimal A No significant increased risk over responses to insects already resident

Benefits Fewer tractor accidents Unlikely Major D At least 3 deaths have occurred while treating CT, see Section 7.2 Significantly fewer skin infections through Improbable Minimal B Benefit likely but of little importance fewer prickles Improved health from reduced Very unlikely Minimal B Benefits minimal when herbicides are used occupational exposure to herbicides safely

6.3 Identification of potential effects on the relationship of Māori and their culture and traditions with their ancestral lands, water, sites, waahi tapu, valued introduced flora and fauna and other taonga (taking into account the principles of the Treaty of Waitangi) Taking account of sections 6(d) and 8 of the Act, list any potential adverse and beneficial effects on the relationship of Māori and their culture and traditions with their ancestral lands, water, sites, waahi tapu, valued introduced flora and fauna and other taonga (taking into account the principles of the Treaty of Waitangi). In this area it is especially important to indicate the extent to which the effects reflect expressed views of the Māori community during your consultations with them. However, details on these views and how they were obtained should be dealt with under the assessment section (section 7).

List risks, costs, and benefits, separately. If you decide that some of them are not potentially significant and do not require further assessment, indicate why you have reached this conclusion.

Risks, costs, and benefits

Key environmental outcomes Potential Potential Comment significant significant risk/cost? benefit? The continued and improved availability, quantity No No The insects will only attack thistles. Decline and quality of traditional food resources (mahinga in thistles might improve the quality of kai) mahinga kai but benefit not significant The continued availability, quantity and quality of No No This might be enhanced by the successful traditional Maori natural resources biological control of thistles but benefit not significant The retention of New Zealand‟s diverse range of No No Thistles are not a significant threat to native indigenous flora and fauna habitats The protection of indigenous flora and fauna No No Thistles are not a significant threat to native valued by Maori habitats The purity of water (inland and coastal) and the No Yes The reduction in herbicide use resulting from need to retain and extend its productive and life- successful biological control could slightly sustaining capacity. reduce pesticide runoff in water, see Section 7 The purity of land and the need to retain and No Yes The reduction in herbicide use resulting from extend its productive and life-sustaining capacity successful biological control could slightly reduce pesticide load on soils, see Section 7. The purity of air and the need to retain and extend No No The reduction in herbicide use resulting from its productive and life-sustaining capacity successful biological control could reduce pesticide drift, but effect not significant The purity of human health and well-being No No The agents will not pose any health risk to humans. Neither agent bites, stings, or causes disease. Successful biological control could reduce human exposure to pesticides but benefit not significant. The restoration and retention of natural habitats No No Thistles are not a significant threat to native habitats

6.4 Identification of potential effects on the market economy Taking particular account of section 6(e) of the Act, list the economic risks, costs and benefits that might arise to New Zealand. Include related effects (e.g. scientific knowledge), which are likely to have economic or related value.

Risks, costs, and benefits

Source of potential risk/benefit Likelihood Severity Risk Comments or score advantage

Market economy

Adverse effects (risks and costs)

Reduced herbicide sales Improbable Minimal B Sales of thistle herbicides not the sole focus of vendors significantly affects vendors‟ businesses Reduced income significantly Improbable Minimal B Thistle management not the sole focus of contractors affects weed management contractors‟ businesses Reduced viability of thistles for Improbable Minimal B Crop and Food Research promotes variegated thistle for essential oil production this purpose. Economic value unproven. VT not a preferred host; agents controllable by insecticides. See Section 7.4 Reduced viability of Cirsium for Improbable Minimal B Current market small, future value doubtful, agents cut flowers controllable by insecticides See Section 7.4 Veterinarians‟ income significantly Improbable Minimal B Thistle-related animal problems not constant, not a reduced because of less callout for significant part of vet incomes scabby mouth Nodding thistle honey production Improbable Minimal B Little effect on market economy but small interests are reduced because fewer thistles vocal. See Section 7.4 Possible non-target effects on Improbable Minor B Exhaustive experimentation suggests no such effect. economically valued plants See Section 7.4 Thistles replaced with a weed that Very Minimal B No other weeds are more damaging is more economically damaging unlikely Reduced flowers lead to Improbable Minimal B Thistle pollen and nectar unlikely to be limiting factor significantly fewer bees for in bee dynamics pollination services and for honey production Reduced thistle control because of Unlikely Minimal Effects expected to be additive. Will be assessed in the competition with other control application agents

Benefits Improved farm production and Likely Moderate E See Section 7.4 earnings from CT control Ditto through incidental control of Likely Moderate E See Section 7.4 other pest thistles Reduced expenditure on thistle Likely Minor E See Section 7.4 control tactics Reduced contamination of frozen Likely Minor E See Section 7.4 peas by Californian thistle buds Reduced costs of scabby mouth Unlikely Minor D See Section 7.4 and other veterinary problems Improved value of wool through Unlikely Minor D See Section 7.4 less contamination Increased income for Landcare Unlikely Minimal C Minor project for the company Research Increased flexibility of farm Unlikely Minimal C management Reduced nuisance value of thistles Unlikely Minimal C in hay Increased profit from fewer pest Improbable Minimal B Thistle seeds influence bird populations uncertain, birds birds in crops because less thistle critical pests of grain rarely flower to sustain populations

6.5 Identification of potential effects on society and communities Taking particular account of section 5(b) and the full definition of “Environment” in section 2 of the Act, list any adverse and beneficial impacts on people and communities that might arise and relate to their capacity to provide for their own social and cultural wellbeing both now and into the future. Also list any ethical or spiritual issues or considerations that might arise as per section 68(1)(a) of the Act. Indicate what steps have been taken to assist the identification of the effects in this area, for example, was there any community involvement, or consultation? However, details on this should be dealt with under the assessment section (section 7).

Risks, costs and benefits:

No cultural, social, ethical or spiritual risks, costs or benefits have been identified apart from those addressed in the table below or elsewhere in Section 6. Details of consultation with Māori can be found in Section 7 and in Appendix A.

Source of potential risk/benefit Likelihood Severity Risk Comments or Score advantage

Society and community

Adverse effects (risks and costs) Rural bird populations significantly Improbable Minor B Assumes bird populations are desirable. Thistle reduced because of fewer seeds for seeds unlikely to be a critical food source for food maintaining populations, and unlikely to be eliminated by biocontrol Significantly increased incidence of Improbable Minimal B Agents will make up only a small part of available wasp stings by wasp populations prey for wasps, no significant increase in wasp increased by eating Cassida larvae populations and hence no significant increase in sting rate predicted Landscape values reduced by decline Improbable Minimal A Thistle patches not highly visible in landscape, and in thistles have minimal or no landscape value, and will not be eliminated by biocontrol

Benefits

Reduced farmer stress from Unlikely Minor D Better control would reduce scabby mouth in sheep. improved animal welfare Animal welfare is important to farmers, but incidence not very common. See Section 7.4 More opportunity for organic Unlikely Minor D CT is an important weed in organic production, but production organics minor in economy. See Section 7.5 Landscape values increased by Unlikely Minimal C Thistle patches not highly visible in landscape but decline in thistles farmers invest in weed control for aesthetic reasons, and in response to neighbour opinion. See Section 7.5 Sustainability of rural communities Highly Minimal B Adverse effects of CT do not threaten community improved improbable stability Reduction in stress in farmers Highly Minimal B May be true, but thistle control not a critical facet of improbable farm management

Ethical issues and considerations: None have been identified.

6.6 Identification of other potential effects (including effects on New Zealand’s international obligations) List any remaining potential adverse and beneficial affects not already covered including any effects on New Zealand‟s international obligations (as per section 6(f) of the Act).

Risks: None identified

Benefits: None identified

Section Seven – Assessment of potentially significant adverse and beneficial effects (risks, costs and benefits)

This section entails detailed assessment of those effects identified in section 6 that you consider to be potentially significant. The assessment should describe the nature of the effects, and should discuss in more detail, than in section 6, the source of the effects and the pathways leading to them. Assessment also entails providing an estimate of the magnitude of the outcome if the effect should occur, and the likelihood of occurrence (which may be measured as frequency or probability). The degree of uncertainty associated with the assessment should also be analysed. The factors set out in clause 33 of the HSNO (Methodology) Order 1998 which outlines various risk characteristics that will influence the decision-makers approach to risk should be referred to. These include characteristics such as the risk will persist over time or the potential adverse effects are irreversible. In such instances the Authority will be more cautious and risk averse when considering such matters.

You should also carry out your assessment taking into account the matters regarding undesirable self-sustaining populations set out in section 37 of the Act (and addressed in section 4 of this form).

ERMA New Zealand uses qualitative scales for assessing effects which may be of some use to you in completing this section – please refer to the ERMA New Zealand Technical Guide “Decision Making: techniques for identifying, assessing and evaluating risks, costs and benefits” for further details. Please cover all of these issues under each of the following headings (areas of impact) that reflect those matters referred to in Part II of the HSNO Act:

7.1 Assessment of potentially significant effects on the environment (in particular on ecosystems and their constituent parts) Assess the potentially significant adverse and beneficial effects associated with the organism(s) to be released and the ways that they might adversely affect or improve/enhance (in the case of benefits) the New Zealand environment e.g. the life supporting capacity of air, water, soil and ecosystems; the sustainability of native and valued introduced flora and fauna; natural habitats and the intrinsic value of ecosystems; New Zealand‟s inherent genetic diversity; animal or plant health.

Assess adverse and beneficial effects (risks and benefits), separately. Where benefits and risks are linked, state this (e.g. a biological control agent which has an impact on both the target organism (benefit) and non-target organisms (risk)).

Risks:

This section assesses the significant direct and indirect risks posed to the environment by the introduction of the two control agents or by reduction in the abundance of Californian thistle as a result of successful biological control. The most significant risk addressed is the possibility that the agents might harm native plant populations. A table outlining the relationship between thistles and other elements of the New Zealand daisy flora is in Appendix A. Other risks covered include direct and indirect adverse effects on native insect species through food web interactions.

Direct effects Risks posed by the control agents to non-target plants Ceratapion onopordi The field host-range records for this species are robust and consistent. In Europe, Ceratapion onopordi has been recorded from a wide range of thistle and related species that belong to the tribe Cardueae. These include Carduus crispus, C. nutans, Cirsium arvense, C. vulgare, Onopordum acanthium, O. illyricum, Centaurea calcitrapa, C.

Briese and Walker (2002), Briese et al. (2003) and Briese (2003) advocate restricting host-range testing to plants that are closely related phylogenetically. Funk et al. (2005) recognise three major clades – Carduoideae, which contains the thistle tribe Cardueae, Cichorioideae, a neighbouring clade that contains the tribe Cichorieae, and Asteroideae, a more distant clade that contains the other tribes represented in New Zealand. As this weevil has not been recorded on plants outside the tribe Cardueae, selection of plants for host-range testing was restricted to the clades Carduoideae and Cichorioideae. In New Zealand the only other tribe within the latter clade that has species of significant environmental, social or economic value is the lettuce tribe Cichorieae. Sonchus kirkii and S. oleraceus, two species valued as food by Māori (puha), were used to represent the tribe Cichorieae in tests. The weevil has never been recorded from sow thistle (S. oleraceus) in Europe. A summary of the relationships between species of the New Zealand daisy family can be found in Appendix A.

Gourlay (2003) conducted laboratory host-range tests and concluded that neither puha species would be at risk from A. onopordi in New Zealand and concluded that host range would be restricted to the tribe Cardueae. However, he found that weevils could lay eggs and complete development on European thistle-like species such as globe artichoke and safflower. As such laboratory tests can overestimate host range, Gourlay (2003) recommended field tests in Europe to elucidate the ecological or true host-range.

The results of subsequent field cage and laboratory tests (Gourlay & Hill 2006; Appendix C) confirmed the hypothesis suggested by field observations: that A. onopordi has a narrow host range, confined to species within the subtribes Carduineae and Centaureinae of the tribe Cardueae. Field tests confirmed that Sonchus species are not hosts for this weevil (Appendix C).

There are few species of potential economic importance in New Zealand that fall within the apparent host range of this weevil. Silybum marianum (variegated thistle) is a potential future crop for essential oil production in New Zealand (although it is also a significant pasture weed in some regions such as Banks Peninsula). This was an acceptable host in field cage tests in Europe (Appendix C), although it is not listed amongst recorded hosts in Europe (Gassman et al. 2005). Centaurea cyanus (cornflower), C. montana (mountain knapweed), and various Centaurea cultivars are grown as ornamentals. Centaurea cyanus successfully supported weevil oviposition and development. Related species are known hosts in Europe, and it is likely that these species will be field hosts in New Zealand. Occasional adults emerged from artichoke (Cynara scolymus) and safflower (Carthamus tinctoria), but use of these species was clearly poor compared with attack on thistle species. Ceratapion onopordi was not recorded as a host of Cynara scolymus and Carthamus tinctoria in Europe by Gassman et al. (2005), and is not recorded as a European pest of either Cynara or Carthamus (CABI Crop Compendium 2004). It is questionable whether incidental adult and larval feeding by A. onopordi would be detrimental to the quality of artichokes and safflower if these become viable crops in New Zealand. The economic consequences of attack on non-target plants of potential economic value are discussed in Section 7.4.

A population of Ceratapion onopordi was held at the Landcare Research containment facility in 2005/06. This offered the opportunity to complete additional tests to reinforce the conclusions outlined above. These tests confirmed that there is no evidence that species of other tribes belonging to the family Asteraceae would be at risk from this weevil in New Zealand. Ceratapion onopordi could not lay eggs or feed on any of the 11 plants selected to represent the other tribes that are indigenous to New Zealand (Gourlay & Hill 2006; Appendix C).

Cassida rubiginosa The field host-range records for this species are also strong. In Europe, Cassida rubiginosa adults have been recorded from several Cirsium and Carduus species, Arctium minus, Silybum marianum, Onopordum acanthium and Centaurea

Cassida rubiginosa became established in North America over 100 years ago. It has been found there on exotic Cirsium and Carduus species, and occasionally on Arctium minus and Centaurea jacea. However, as in Europe, there are no records of it occurring on plants outside the tribe Cardueae in its adopted range. For this reason, recent host- range testing was restricted to plants belonging to this tribe.

As with Ceratapion onopordi, green thistle beetle has not been recorded on plants outside the tribe Cardueae. Selection of plants for host-range testing was restricted to the clades Carduoideae (to which the tribe Cardueae belongs) and the neighbouring clade Cichorioideae (Funk et al. 2005). Sonchus kirkii and S. oleraceus, two species valued as food by Māori (puha), were used to represent the tribe Cichorieae in tests. A summary of the relationships between species of the New Zealand daisy family can be found in Appendix A.

Testing suggested that the host range of the green thistle beetle is slightly wider than that of Ceratapion onopordi, although still restricted to species within the tribe Cardueae. Most if not all species in subtribe Carduinae, and perhaps most in subtribe Centaureinae, appear to be suitable host plants for oviposition and larval development of C. rubiginosa in the field in Europe. In European field cage tests minor adult feeding and a very small number of eggs were recorded on plants in the subtribes Echinopsidinae, Carlininae, and Centaureinae. Larval starvation tests showed that larvae were capable of developing to adult on these plants. Centaurea cyanus supported full development of 80% of larvae tested, and adults could lay eggs on this plant. These results suggest that plants in this subtribe, including safflower (Carthamus tinctoria), would be capable of supporting populations of green thistle beetles once established in New Zealand. The green thistle beetle did lay eggs on Sonchus oleraceus (exotic „puha‟), but no larval feeding or development occurred on this species. This suggests S. oleraceus is not capable of supporting a colony of the green thistle beetles in New Zealand (also see Section 7.3).

Because New Zealand has no natives in the tribe Cardueae and almost all imported thistles and knapweeds in this tribe have the potential to become weedy once established in the New Zealand environment, it would be beneficial to have biocontrol agents that are oligophagus within the tribe. This would mean that any thistle species is likely to be attacked by both the weevil and the beetle while leaving any New Zealand native plant species intact.

Competition with native insects The insect fauna associated with Californian thistle in New Zealand is poorly known. There are no species known that would suffer adversely in competition with these agents.

Loss of existing control of thistles in native habitats Thistles are not a significant component of native habitats, and so any loss of existing biological control of thistles would have no significant effect on the integrity of those habitats. This issue is discussed further in Section 7.4.

Indirect effects Potential adverse effects on native flora and fauna of introducing two new insect species The establishment of Ceratapion onopordi and Cassida rubiginosa would introduce a new element to the ecosystems surrounding thistle plants and populations. The introduced species would add to the biomass of insects in the environment, and these would be available to parasitoids and predators. Food webs would be altered if changes in the abundance of resident natural enemies generated by the addition of the two beetle species had a significant adverse effect on populations of native insects, or on the animals or plants on which they feed. Is this likely?

Thistles are a species typical of pastures, grasslands, and cultivated land, and rarely impinge significantly on native habitats except in native grasslands, riverbeds, river margins, and amenity areas (K. McAlpine, DOC, pers. comm.;

Appendix A). This means that there is only limited opportunity for thistles or their natural enemies to interact with the flora and fauna of native habitats. For this reason alone, the risk of the two agents significantly affecting food webs in native habitats is considered to be insignificant.

Eggs of Ceratapion onopordi would be available to those egg parasitoids and predators that frequent pastures in New Zealand (B. Barratt, AgResearch, pers. comm.; Appendix A). Larvae and pupae of C. onopordi occur inside stems and are probably unavailable to generalist parasitoids and predators. Adult weevils would be available to spiders. However, weevils are small and will be relatively uncommon. It is unlikely that the addition of this weevil will add significantly to the prey biomass in the pasture ecosystem as a whole, and so indirect effects on other prey species or plants are likely to be minimal (B. Barratt, AgResearch, pers. comm.; Appendix A).

Eggs of Cassida rubiginosa would be available to those egg parasitoids and predators that frequent pastures in New Zealand (B. Barratt, AgResearch, pers. comm.; Appendix A). The larvae and pupae of Cassida rubiginosa are likely to be abundant on thistle leaves. These larvae may be attacked by wasps such as paper wasp and German wasp. If this increased the biomass of wasps then the predator pressure on other species present, such as flies, may increase.

As described in Section 3.4, the parasitic wasp Eupelmus vesicularis (Retzius) is responsible for only 0.1% of the parasitism of Cassida rubiginosa in northern Virginia. This parasitoid is present in New Zealand. If C. rubiginosa is introduced as a control agent it may be attacked by E. vesicularis. There is not enough known about the ecology of this parasitoid and its host species in New Zealand to predict the effect of introducing a new host species into existing food webs (e.g. Jo Berry, Landcare Research, B. Barratt, AgResearch, pers. comm.; Appendix A). This species has a very wide host range, and it is unlikely that the introduction of one more host will significantly perturb existing host– parasitoid relationships.

As thistles are not closely associated with native habitats, any impact on food webs is likely to be limited to managed grasslands, and the risk of indirect adverse effects on non-target species in native habitats is considered insignificant. In grasslands, Californian thistle forms monocultures, and can form a dominant but patchy component of the plant biomass at a paddock level. Any effects on food webs are likely to be restricted to the vicinity of these patches, to be local, ephemeral, and reversible.

Indirect effects of reducing the incidence of Californian thistle in the environment These effects are likely to be insignificant. Successful biological control could reduce the amount of Californian (and other) thistle in grasslands. Increased higher quality forage, and reduced risk could lead to increased stocking rate, increasing grazing pressure, potentially increasing faecal matter, and decreasing water quality. Although this effect may occur on occasions, the magnitude of the possible effect is minimal and the risk low.

Less Californian thistle would reduce the pool of pollen and nectar available for native insects and honeybees. This thistle is abundant only locally, and provides a relatively small proportion of the total available pollen and nectar available at a landscape level. Californian thistle would be replaced in part by pollen- and nectar-bearing plants such as clover. The effect is not unlikely, but the size of the effect on insect numbers is likely to be minimal. The risk is regarded as low.

Reducing the abundance of Californian thistle could reduce the abundance of these insect species and hence faunal biodiversity in pastoral ecosystems. The native insect fauna of thistles is poorly known, so the effect of this is uncertain. However, even under successful biological control Californian thistle is likely to be a common component of grassland flora, and the risk of faunal loss is considered insignificant. There are no insect species known that will be displaced by either Ceratapion onopordi or Cassida rubiginosa.

Potential adverse effects of replacement of Californian thistle by other weeds Californian thistle is not a significant component of native habitats, and so there are no significant risks posed from replacement in such habitats.

Benefits: The staff of the Department of Conservation (DOC) do not see Californian thistle as a major conservation issue, but there are examples where it competes with native flora and adversely affects natural processes (K. McAlpine, DOC, pers. comm.; Appendix A). It limits the success of restoration of native vegetation on Quail Island. Successful biological control would reduce the competitive ability of Californian thistle in low-stature native habitats. These effects are locally significant but uncommon, and the overall benefits of successful biological control are likely to be minor.

DOC conservancies apply herbicides to Californian thistle, but clearly the scope of control measures is small, and any projected benefits from reducing herbicide use would also be minor (K. McAlpine, DOC, pers. comm.; Appendix A).

The indirect effect of the introduction of these two insects on food webs is potentially complex and difficult to assess (see above). Any impacts on biodiversity values in grassland ecosystems are equally likely to be beneficial or adverse.

7.2 Assessment of potentially significant effects on human health and safety (including occupational exposure) Assess any potentially significant adverse and beneficial effects on human health that may be related to the release of the organism(s) in New Zealand. If effects in this area are likely to be significant a full health assessment as set out in the relevant ERMA New Zealand technical guide may be warranted.

Assess potentially significant adverse and beneficial effects, separately.

Risks:

No significant risks have been identified (Section 6.2).

A search of “Pubmed” offered no relevant references.

Benefits:

Labes (2000) noted that at least three tractor drivers have died while undertaking operations associated with thistle control. While successful biological control might eliminate the possibility of such accidents, so might safety education. No claims are therefore made for the benefits of biological control in this area.

Successful control of Californian thistle would eventually reduce the amount of herbicide applied to infestations in New Zealand, and so reduce the frequency of exposure of the public and spray operators to the toxic elements of herbicides. Assuming that herbicides are currently applied according to label requirements and in accordance with requirements for occupational safety, these benefits to human health are presumably marginal.

7.3 Assessment of potentially significant effects on the relationship of Māori and their culture and traditions with their ancestral lands, water, sites, waahi tapu, valued flora and fauna and other taonga (taking into account the principles of the Treaty of Waitangi) Assess the potentially significant adverse and beneficial effects on the relationship of Māori and their culture and traditions with their ancestral lands, water, sites, waahi tapu, valued flora and fauna and other taonga (taking into account the principles of the Treaty of Waitangi). If there are potentially non-negligible effects to consider in this area, it is expected that consultation will have occurred with Māori, and any information obtained from this consultation should be appended to the application. Give details of this in the space provided (see the User Guide for what is required).

Assess potentially significant adverse and beneficial effects, separately.

Consultation with Māori:

A letter seeking consultation, accompanied by an „information pack‟, was circulated to a list of 90 iwi, hapū, rūnanga, and Māori organisations supplied by ERMA NZ. This consultation document was distributed in early July. Responses were requested by 28 August (6 weeks) but were accepted until mid-October.

The distributed material is reproduced in Appendix A. It described how the applicant intended to assess the risks, costs and benefits surrounding the proposed introduction of Ceratapion onopordi and Cassida rubiginosa in the application, and asked each organisation to identify any issues that were inadequately or not covered in those plans. Information sheets about safe practice of biological control of weeds were included, along with answers to „frequently asked questions‟. Recipients were given the option of responding by form letter (a self-addressed envelope was included), by email, or by phone, and were invited to seek more interaction with the applicants. The offer to meet kanohi ki te kanohi was not explicit, but such meetings would have been undertaken had respondents requested it.

Email or written responses were received from 12 sources including one representing the 18 rūnanga of Ngāi Tahu. Comment was therefore obtained from 30% of correspondents. The applicant entered into dialogue on all specific issues raised by respondents as and when requested. In none of the written responses, or in the follow-up phone calls, was a specific request made for a face-to-face meeting.

Most respondents generally felt that the information provided in the information pack either allayed concerns they held or gave them confidence that the ERMA process would be adequate to address their concerns. The issues raised are detailed in Appendix A, and addressed in this section. One respondent requested that the applicant provide resources to facilitate consultation locally. This was outside the approved budget of the applicant and was refused. Most concerns raised by respondents were assuaged by further discussion. This included one respondent who noted that there was limited information provided, insufficient time available to complete consultation, limited resource available to complete consultation, and that there was no prescribed face-to-face consultation. The issues raised during consultation are detailed in Appendix A and are addressed below.

Once the application is lodged the applicant intends to inform all organisations how to access the application and prepare submissions.

Risks:

The key issues identified by Māori correspondents during this consultation process and in dialogue over previous similar applications are as follows:

Potential impacts on native daisies, puha, or other native species The cultural importance of puha (Sonchus species) was emphasised by several respondents (Appendix A). The applicants acknowledged this sensitivity when planning host-range tests for both agents. Seeds of Sonchus kirkii and S. oleraceus were grown in Europe, and their susceptibility to both agents was measured in field tests. The results of these and other tests are presented in Appendix C and summarised in Section 7.1. Neither species is at risk from Ceratapion onopordi or Cassida rubiginosa.

Scientifically sound and internationally accepted protocols for selecting a representative panel of test plants have been established (Wapshere 1974) and were used in this case. The concerns raised by Ngāi Tahu about „functional similarity‟ between plants (Appendix A) are taken into account when host-plants are selected for testing using these protocols. There is growing confidence amongst researchers that relatedness alone is an adequate indicator of the relative risk posed to non-target plants by specialist control agents and that the numbers of test plant species requiring testing could be reduced, and that plant structure, biochemistry and seasonality need only be considered within the relatedness framework (Briese & Walker 2002; Briese 2003).

The reasoning behind the selection of plants to be tested against Ceratapion onopordi and Cassida rubiginosa can be found in Appendix C and in Section 7.1. The relationship between thistles and native New Zealand daisies is summarised in Appendix A. The results of tests are summarised in Section 7.1.

Adverse effects of new organisms on existing faunal relationships Risk of attack to non-target native plants was a primary concern for all respondents. Host-range testing indicates that these insects will not feed on plants outside the thistle tribe, and no native plants are at risk of attack. The evidence for this conclusion can be found in Appendix C, and the conclusions are summarised in Section 7.1.

Monitoring Te Rūnanga o Ngāi Tahu advocated monitoring of population build-up and dispersal, effectiveness of control, and non- target impacts (Appendix A).

The simple programme planned to monitor the establishment and spread, and to identify obvious effects on non-target native plants growing in release sites, is presented in Section 5. The applicants consider this to be the appropriate level of research for this stage of the project.

It is always desirable to measure the success or otherwise of biocontrol agents, but this process is difficult to plan. Once released, agents usually persist at low density for some years, and can barely be detected. Population build-up to the point of equilibrium density (and therefore maximum impact) usually takes many years, and it is not valid to conduct field research to define the impact that the agent will have until this equilibrium density is reached. Just how many years this will take is rarely predictable. Investment in detailed monitoring (even of population build-up and dispersal) at this stage of development is not justifiable scientifically or economically.

Reliable measurement of agent impact is long-term, time-consuming and expensive, and is beyond the planning horizon of most science funding bodies (which are not usually interested in funding this style of research anyway). This research is certainly beyond the present resources and scope of the applicant.

The presence of non-target damage on non-host plants is the simplest measure of non-target effects, and such observations will be made at release sites. However, absence of such damage may simply reflect the rarity of agents in the early years of establishment, and more reliable assessment of such effects must also wait populations‟ build-up. The same argument applies to measuring the impact of agents on established food webs. The difficulties here are compounded by the almost total lack of information about the „steady state‟ of such webs in New Zealand against which to measure change.

Predicting future agent behaviour Respondents expressed the view that new organisms should not be introduced because it is impossible to predict future risks to the environment and especially to native species (Appendix A). There are two issues here. The population growth of introduced species is governed by the effects of any resident natural enemies the new organisms might attract, and how developmental and reproductive rates of the new organisms react to the new climate. For these reasons, the applicants acknowledge that whether a species establishes in New Zealand, how quickly populations build up, and the final equilibrium population density likely to be achieved are questions that are difficult to predict. However, it is important to distinguish these population processes from the physiological processes that govern host plant selection by individual insects. These behaviours are governed by genetics and are only amenable to evolutionary change, not changes in climate or environment. Sound knowledge about these host selection processes are gained through host-range-testing experiments, and these results can be used reliably between countries and between climates. There are no examples known of weed control agents changing host range between climates.

Biological control and integrated pest management Biological control is not promoted as a „one-stop‟ alternative to other forms of weed management. Should these agents prove to be so effective that thistles are no longer a problem to pastoral farmers anywhere, then alternative (and less desirable) management techniques such as herbicide application will become redundant. However, it is more likely that control will establish gradually, and effectiveness will vary from place to place and from time to time. Other control methods will still be required. The benefits of successful biological control will be to reduce the area treated or the frequency of treatment, or to allow the use of more environmentally friendly methods. The level of benefit that would result from integrating biological control and more conventional management techniques is difficult to predict (Section 7.4).

Consultation with local iwi before first release in any area This is a generic issue relevant to all weed biological control projects, and has been taken up with regional councils and Landcare Research.

Benefits:

No potential benefits of the proposed introductions were identified by Māori in pre-application consultation because the primary aim of consultation was to identify risks and costs.

The benefits detailed in other parts of Section 7 apply to all.

As 2.8% of the total area of sheep/beef, sheep, beef, and dairy farmland defined as „grassland‟ in 2002 (not including tussock and Danthonia for grazing) belongs to Māori Incorporations and Trusts (Appendix B) it is assumed that at least 2.8% of all benefits gained from improved production in grassland will be captured exclusively by Māori (http://www.maf.govt.nz/statistics/primary-industries/index.htm). These benefits may be substantial (see Section 7.4).

Successful biological control of Californian thistle would eventually lead to reduction in the use of herbicides. Reduced herbicide application would reduce the risk of contamination of water and land, issues of great importance to

Māori. Assuming that herbicides are used according to label at present, the rate of contamination is currently low, and therefore benefits will not be great.

7.4 Assessment of potentially significant effects on the market economy Assess the potentially significant adverse and beneficial effects on the market economy. Effects on third parties and to New Zealand of the proposed release need to be specifically evaluated. If economic effects are significant applicants should provide a cost benefit analysis. Guidance on the requirements for cost benefit analyses are set out in the ERMA New Zealand Technical Guide on Economic risks.

In this case it is still helpful to assess risks and costs, and benefits, separately but if possible these assessments may be drawn together into an overall cost benefit analysis. As a part of this, estimate net benefits.

Risks:

The market economy could be adversely affected if: Ceratapion onopordi and Cassida rubiginosa directly interfered with agricultural production Reduction in thistle populations reduced agricultural production or affected the viability of the rural sector Competition between control agents reduced current control of nodding thistle Californian thistle was replaced by a more damaging weed.

Costs of non-target damage to valued crops Host-range testing has shown that the risk of either Ceratapion onopordi or Cassida rubiginosa causing significant damage to plant species outside the tribe Cardueae is insignificant (see Section 7.1). However, there are several species within the tribe that have potential economic value in New Zealand. The current value of the crops is insignificant and so the magnitude of any risk to the market economy is insignificant. The likelihood of risks to these plants is as follows:

Globe artichoke (Cynara scolymus) – Two Ceratapion onopordi adults emerged in field cage tests (compared with 87– 201 from Carduus species). Eggs were laid on globe artichoke plants in field cages, although far fewer than on other thistle-like species. Hatching larvae could develop to adulthood in the laboratory. Despite these experimental records, neither insect has been recorded as a pest of globe artichoke in Europe, where it is a major crop. (CABI Crop Compendium 2004, http://www.inra.fr/Internet/Produits/HYPPZ/CULTURES/3c---035.htm). It does not appear to have been recorded on Cynara scolymus in North America.

Globe artichoke is grown occasionally in home gardens in New Zealand, and there is one commercial grower in Pukekohe (although it is being promoted as a future crop; Crop and Food Research 2001). Given European experience, the risk of incidental attack by either species on globe artichoke in New Zealand is considered to be unlikely. There is no risk of significant damage to the harvestable head. Any significant damage to leaves or stems might be mitigated by insecticide application.

Safflower (Carthamus tinctoria) – Only three Cassida rubiginosa egg masses were laid on safflower in field cages (compared with 57 on Californian thistle), and survival to adult in feeding tests was poor. Only 17 Ceratapion onopordi adults emerged from this plant in field cages (compared with 87–201 from Carduus species). Despite these experimental observations, neither insect has been recorded as a pest of safflower in Europe.

Safflower is grown for oil production elsewhere in the world, but there does not appear to be any current commercial interest in New Zealand for oil production. This plant is not a known host of either species in Europe, and the risk of

Variegated thistle (Silybum marianum) – Both agents completed development on this plant, although development success was less than on other thistle species. It is possible that it will be a suboptimal field host for both species in New Zealand. The risk of incidental feeding by both species on variegated thistle is likely.

Silybum marianum is grown commercially for seed in Europe, Egypt, China and Argentina as a source of silymarin, an extract used to treat liver complaints. Trials have been conducted in New Zealand since 1994 to assess the potential for an industry here (Martin et al. 2006), but there appear to be no commercial plantings at present. Risk to future commercialisation of variegated thistle might be mitigated by insecticide treatment.

Cornflower, other Centaurea species – It is clear that Centaurea species are adequate hosts for both agents, and are recorded hosts for both species in Europe. The risk of both species attacking Centaurea species on release in New Zealand is considered likely.

Many Centaurea species are notorious weeds worldwide, and some of these are naturalised in New Zealand. However, cornflower (Centaurea cyanus) and other species are commonly grown as ornamentals, and would be at risk from these agents. They could be replaced in gardens by non-susceptible species, or the effects of the insects could be mitigated by insecticide application.

Indirect costs resulting from reduced incidence of Californian and other thistles The net indirect costs of successful control of Californian thistle are considered to be minimal (Section 6.4). Successful control might result in less work for vets attending cases of scabby mouth and undertaking vaccination programmes, but this is unlikely to be a significant proportion of a vet‟s business. Successful control could reduce the work available to agricultural contractors and helicopter operators in thistle-prone areas. However, most on-farm thistle control is carried out by farmers and their staff (Moller et al. 2006; Appendix B) and so this effect is not likely to be large. Herbicide sales by suppliers might also be reduced.

If control of thistles was effective, less thistle pollen would be available for honeybees to harvest to feed larvae in the hive. However, thistles flower in January and February when many alternative pollen sources are available, and when the demand for pollen in hives is not high (B. Donovan, beekeeper, pers. comm.). The reduction in pollen available to bees from thistles would not be a significant cost to beekeeping. Reduction in the incidence of nodding thistle might reduce the numbers of nectar-bearing flowers available to honeybees and reduce the production of nodding thistle honey. Thistles are likely to be replaced in part by nectar-bearing plants such as clover.

Potential loss of existing level of thistle control provided by other control agents Other biocontrol agents have been introduced to attack thistles. The agents previously introduced to attack Californian thistle either did not establish, or are present in numbers too small to cause significant damage to the thistle (Section 3.4). For this reason introduction of Ceratapion onopordi and Cassida rubiginosa would not reduce the current level of biological control of Californian thistle.

Biological control programmes have also been instituted against nodding thistle and Scotch thistle, both of which will be hosts for the two new agents. Models suggest that nodding thistle crown weevil contributes to nodding thistle control (Shea and Kelly 2004) and anecdotal information suggests that a significant level of control has been achieved in New Zealand. Both Ceratapion onopordi and Cassida rubiginosa could compete directly with this weevil by feeding on and destroying rosettes. Both could compete indirectly with the nodding thistle receptacle weevil and gall fly, and with the Scotch thistle gall fly by reducing the abundance of flowering stems.

The literature provides few examples anywhere of asymmetric competition, i.e. competition that strongly benefits one antagonist over another. In a recent review, Reitz and Trumble (2002) found only 42 cases of competitive displacement in the entomological literature, and only one involving insects feeding within a plant. Biological control of weeds normally involves the introduction of more than one control agent (Julien & Griffiths 1998). There are no examples of projects where the introduction of subsequent agents has reduced the overall level of control achieved. The applicants therefore conclude that even if competition did occur, the existing level of control of nodding thistle and Scotch thistle would not be reduced. Interactions are likely to vary from place to place and from time to time.

Risk of replacement by a more damaging weed Californian thistle is regarded as the most important and persistent herbaceous weed problem in New Zealand pastures (Rahman & Popay 2001; Popay et al. 2002; Moller et al. 2006; Appendix B). It is unlikely that any other herbaceous weed would occupy pasture so comprehensively, and so adverse effects would probably be reduced. Should successful control be achieved, the mostly likely species to replace Californian thistle are perennial fodder plants such as ryegrass and clover, which already grow between thistle stems.

The wider host range of these agents could prevent the invasion of other thistle species that threatened to replace Californian thistle.

Benefits:

Californian thistle reduces the productivity of pastures mainly by deterring grazing animals from walking between the prickly stems and grazing the fodder growing there. This usually means that the pasture area inside the margins of the infestation is completely lost to grazing by sheep. Californian thistle also reduces the growth of pasture plants by shading. It is an aggressive weed that must be managed to maintain the productivity of infested pastures.

The market economy would benefit from biological control of Californian thistle if: The annual cost to farmers of controlling Californian thistle was reduced Stocking rates and farm incomes increased because more pasture was available to grazing animals Animals were healthier Crops grew free of competition from Californian thistle. The benefits estimated below would accrue if biological control was successfully managed by biological control alone.

Reducing the cost of Californian thistle control on South Island livestock farms Labes (2000; Appendix B) found that Californian thistle reduced pasture production, increased the likelihood and severity of the viral infection „scabby mouth‟ in lambs, reduced the quality of hay and winter feed, and affected success of grain crops. The farmers estimated average expenditure on chemical control at $1,200 annually, but most farmers also topped (mowed) their thistles. The survey concluded that farmers in the Otago/Southland Region probably spent $6.6 million on chemicals, $18 million on topping, and $2.4 million on „scabby mouth‟ vaccine annually. The current value of Labes‟ (2000) estimate of the cost of Californian thistle control in Southland and Otago is $32.2 million per year (Appendix B).

Moller et al (2006; Appendix B) conducted a survey of weed management practices on selected sheep/beef farms throughout the South Island. They found that Californian thistle was the most important herbaceous weed on these farms, and expenditure on control exceeded that on higher profile weeds such as gorse and broom (Appendix B). Unlike earlier studies (Mitchell & Abernethy 1993; Labes 2001), this survey explicitly partitioned the benefits of topping for pasture improvement from topping for thistle control. Farmers in this survey spent an average of $2,381 per annum on Californian thistle control or $5.16 per hectare.

MAF land-use statistics state that as at June 2002 there were 240 sheep and beef cattle farms in Otago and Southland, and 700 in the South Island. The Californian thistle control costs obtained by Moller et al. (2006) were obtained from sheep and beef cattle farms only. The annual cost of Californian thistle control ($2,381) to all farms of this type alone in Otago and Southland is therefore estimated at $0.57 million, and in the South Island at $1.7 million.

Table. Estimated annual management costs and production losses attributable to Californian thistle on sheep and beef cattle farms and livestock and grain farms in Otago/Southland, and in the South Island ($000s). The assumptions behind the estimates can be found in Appendix B. The estimates are in 2006 dollars.

Annual estimate ($ 000s) derived from

Measure of management cost or Mitchell & Van Toor & Labes 2000 ARGOS/ Moller production lost Abernethy 1993 Stuck 1993 et al. 2006

Sheep/beef cattle farms Otago/Southland Control costs 570 Production lost 2,300 Combined 2,870

Livestock and grain farms Otago/Southland Control costs 32,200 16,800 Production lost 14,100 67,000 Combined 32,100 83,800

Sheep/beef farms South Island Control costs 1,700 Production lost 6,600 Combined 8,300

Livestock farms South Island Control costs 41,500

This survey did not estimate the costs of managing Californian thistle infestation on other types of livestock farms (Moller et al. 2006; Appendix B), but assuming the same average expenditure, the current annual expenditure on Californian thistle management on all 7035 such farms in Otago and Southland would be $16.8 million (compared to Labes‟s (2000) estimate of $32.2 million), and across the South Island would be $41.5 million.

Reducing loss of pastoral productivity to Californian thistle on South Island livestock farms Mitchell and Abernethy (1993) conducted a survey of farmers on intensely grazed pasture land in southern Otago and Southland. They found that on average 33% of the grazing area on each farm carried Californian thistle infestations. The current value of their estimate of the annual cost of control and loss of production caused by the weed in Otago and Southland was $32.1 million. Van Toor and Stuck (1993) also surveyed farmers in these regions to seek information from which to calculate production losses. Californian thistle was the most important weed to Central Southland farmers, with losses estimated to be $3,101 per farm. Barley grass, gorse and nodding thistle were all considered more

The ARGOS project found that an average of 2.8% of sheep/beef „grassland‟ pastures surveyed in the South Island was occupied by Californian thistle (Appendix B). Assuming that loss in gross revenue was proportional to loss of grazing, then losses to sheep and beef cattle farming attributable to Californian thistles in the South Island alone is estimated at $6.6 million annually.

Assuming that the effects of Californian thistle are the same in southern regions as in the rest of the South Island, then annual production loss on „grassland‟ on sheep/beef cattle farms alone in Otago and Southland is estimated to be $2.3 million.

There are 7035 livestock farms in Otago and Southland, all of which are more or less susceptible to Californian thistle. Sheep/beef farms make up only 3.4% of those farms (sheep farms make up 60% of farms, and dairy farms 13%; MAF Statistics). If sheep and beef cattle farms suffer losses of 2.3 million, and if the same quantum of losses can be attributed to all farms, then annual losses to Californian thistle across all livestock and grain farms in Otago and Southland could be as high as $67 million.

Reducing the costs of Californian thistle elsewhere in New Zealand Hackwell and Bertram (1999) considered that the estimates of Mitchell and Abernethy (1993) represented a degree of double-counting of production losses and control costs, and arbitrarily discounted by 50% the estimated $21,000,000 cost of Californian thistle to the Otago/Southland region. However, these components are not mutually exclusive. The survey of Van Toor and Stuck (1993) demonstrated that real production losses occurred on Central Southland farms despite significant contemporaneous expenditure on weed control. Labes (2000) pointed out that control measures against Californian thistle are often aimed at stopping Californian thistle infestations from getting worse rather than mitigating their effects on production.

Hackwell and Bertram (1999) noted that it was difficult to apply an appropriate multiplier across the country because of a paucity of information on the incidence of Californian thistle and its effects on agricultural production elsewhere in New Zealand. They therefore used the conservative figure of $10 million to represent the effect of Californian thistle on New Zealand agriculture. The information presented in Appendix B suggests that this estimate is too low. Bourdôt and Kelly (1986) measured ground cover of Californian thistle in randomly selected paddocks in three regions (two in the North Island one in the South Island) over two years. Average cover ranged from 0.06 to 3.39%. In a recent survey of monitor paddocks on 30 sheep/beef farms in the South Island (ARGOS project, unpublished data; Appendix B), found that Californian thistle ground cover averaged at least 2.8%. Rahman and Popay (2001) found that thistles (of which Californian thistle was the dominant species) were considered as abundant and as serious in the North Island as in the South Island. Although data are limited, it is therefore not unlikely that Californian thistle occupies at least 1% of New Zealand pastures. With the market value of products from New Zealand pastures exceeding $5 billion, the opportunity cost of grazing lost to Californian thistle is likely to be large.

„Scabby mouth‟ is a parapox virus that infects sheep, especially lambs. It infects scratches and abrasions, often around the mouth where sheep have been feeding on thistles. Infected sheep find grazing difficult for several weeks, with consequent loss of condition. Mortality can reach 5% (Appendix B). Treatment is rarely successful, and so grazing management and vaccination provide the best means of control. The value of reduced scabby mouth infection has not been captured here.

Reducing the costs to the horticultural industry Californian thistle is a major problem in a variety of vegetable crops but there is insufficient hard data available to estimate the current annual cost of ameliorating those effects across the industry (S. Whiteman, Horticulture New Zealand, pers. comm.). It is a particular problem in organic crops such as asparagus where mowing or constant cultivation is one of the few options available to growers (Teulon et al. 2005). Heinz Wattie estimates that the weed is responsible for approximately 1-2% of the total cost of production in their vining pea crops, and extrapolated across the New Zealand process pea industry alone equates to a cost of $0.5–1 million (B. Snowden, Heinz Wattie, Appendix A). Californian thistle is also an issue for arable farmers.

Overall cost benefit analysis:

Also see section 7.7

The costs of this proposal to the market economy are considered to be minor. The cost of completing the project is small. Costs to rural services of successful control would be relatively small, and most costs represent transfers within the rural economy. If biological control of Californian thistle is likely then occasional damage to garden centaurea species and variegated thistle crops is possible. The current value of the thistle crop is insignificant. Garden plants could be protected by insecticide, or other plants could replace them at minimal cost.

The potential benefits of successful biological control of Californian thistle presented here are considered conservative as the costs of thistle infestation to most categories of agricultural production have not been estimated. Californian thistle is a serious weed in most agricultural, and in many horticultural areas in New Zealand. For example, the area of New Zealand under sheep, cattle and dairy farming is over 10 times that under sheep/beef farming. Given the estimates for production losses associated with Californian thistle in sheep/beef farming presented above, the total losses to New Zealand agriculture must be very large. However, there is insufficient information about the following farming systems to calculate a reliable estimate of the national production losses (or control costs) imposed by Californian thistle: Sheep farming Dairy farming Beef cattle farming Grain-livestock farming Vegetable growing Grain growing Other horticulture Other livestock farming

The estimates of potential loss to Californian thistle presented here are also considered conservative because the monetary benefits of biological control accruing from the following are not included: Reduction in control costs and increased production in Otago and Southland from control of other thistles Benefits to New Zealand pastoral agriculture (outside Otago and Southland, and South Island sheep/beef farms) from biological control of thistles Reduction in the cost associated with avoiding contamination in organic pea production Reduction in small but non-zero control costs currently incurred by DOC and local authorities Increased pastoral production from reduced collateral herbicide damage to nitrogen-fixing clovers Reduction in farm costs associated with treatment and vaccination for „scabby mouth‟.

Both control agents attack other thistles. Moller et al. (2006) found that nodding thistle and Scotch thistle ranked third and fourth (after Californian thistle and gorse) in the list of weeds treated by sheep and beef cattle farmers in the South

Island in the past five years (Appendix B). No account has been taken of the economic benefits that would accrue to New Zealand agriculture if improved biological control reduced the impact of other thistles on pastoral production.

7.5 Assessment of potentially significant effects on society and communities Assess the magnitude and distribution of any adverse and beneficial impacts on people and communities that adversely affect or maintain/enhance (in the case of beneficial impacts) their capacity to provide for their own social and cultural well-being both now and into the future. Also discuss any ethical or spiritual issues or considerations that might arise. If social effects in particular (although it may apply to other effects also) are likely to be significant, a full impact assessment may need to be carried out. Refer to the relevant ERMA New Zealand Technical Guides for assistance. If community consultation has been carried out to assist the assessment, provide information on how this was done and the results.

Assess potentially significant adverse and beneficial effects, separately.

Community Consultation:

CalTAG is a community-based initiative originating in Southland and Otago. It is supported by funding or in-kind support from Environment Southland, Horizons Manawatu-Wanganui, Clutha Agricultural Development Board, FRST, Foundation for Arable Research, Landcare Research, AgResearch, and Meat and Wool Innovation. The Group was formed in 1999.

Van Toor and Stuck 1993) consulted farmers in Southland and Otago in the preparation of their cost–benefit analysis. Moller et al. (2006) consulted sheep and beef farmers in the South Island, and Popay et al. (2002) consulted the farming community, local authorities, government departments and farming bodies when assessing the relative importance of Californian thistle to hill country farmers.

In preparing this application, an information pack was assembled, and distributed to 90 iwi organisations (see Section 7.3), Federated Farmers of New Zealand, Foundation for Arable Research, New Zealand Institute of Crop and Food Research, Heinz Wattie, ENZOL, Horticulture New Zealand, and the Nursery and Garden Industry Association. The pack is reproduced in Appendix A, along with the full list of those consulted and their responses. Opinion was also sought from selected regional council biosecurity officers. The Department of Conservation was consulted. All issues raised in the course of this pre-application consultation are addressed in this application. Personal communications from other correspondents are acknowledged in the text and in Appendix A.

Risks:

Apart from the issues already addressed in Sections 6 and 7, no additional cultural, social, ethical or spiritual risks or costs have been identified.

Benefits:

ARGOS (unpublished data) found that animal welfare was a major pre-occupation of sheep and beef cattle farmers in the South Island. Farmer stress would be reduced if biological control of Californian thistle reduced the incidence of „scabby mouth‟ in sheep. Apart from the issues already addressed in Sections 6 and 7, no additional cultural, social, ethical or spiritual benefits have been identified.

Ethical issues and considerations:

None identified

7.6 Assessment of other potentially significant effects (including New Zealand’s international obligations) Assess any remaining potentially significant adverse and beneficial effects not already covered including any effects on New Zealand‟s international obligations. Specify any relevant international agreements.

Assess potentially significant adverse and beneficial effects, separately.

Risks and costs:

No additional issues have been identified

Benefits:

No additional benefits have been identified

7.7 Overall evaluation of adverse and beneficial effects (risks, costs and benefits) It is the role of the Authority to decide whether the beneficial effects (benefits) of the release outweigh the adverse effects (risks and costs). However, if you have a view on the relative importance of the different risks, costs and benefits and how they should be brought together in the overall evaluation of your application then please state that here.

The cost of the project The cost already incurred by CalTAG in developing the biocontrol agents to this point is $395,000.

The additional cost of importing, completing quarantine, and establishing each agent in at least one site in Otago or Southland is estimated to be $60,000. If 40 additional founding populations were established across New Zealand, the lag time between initial release and the maximisation of benefits would be significantly reduced. The additional cost would be $80,000. No other direct economic costs associated with introduction are anticipated. It is assumed that both agents will establish on release.

Likelihood of control Successful biological control of Californian thistle would be achieved if the density and coverage of stems was reduced sufficiently for animals to resume grazing efficiently, and if control measures became less common because fresh invasion of pasture was balanced by patch regression as a result of insect attack. Ang et al. (1995) showed that a combination of plant competition and defoliatioin by Cassida rubiginosa led to stem mortality and could aid control. Friedli and Bacher (2001) suggested that Ceratapion onopordi increased the effects of the systemic rust fungus Puccinia punctiformis (which is already present in New Zealand). They showed that the synergistic interactions between insects, disease and competing plants reduced root biomass by 75% over 2 seasons, and considered the outlook for biological control of this weed to be promising.

Korte et al. (in Labes 2000) suggested that control of Californian thistle could be considered successful if the abundance of stems was reduced to 0.5 stems/ m², especially if the height and vigour of those stems was reduced. Successful control does not require Californian thistle stands to be destroyed, and for this reason the probability of achieving 50% reduction of the costs of Californian thistle to the market economy are estimated to be 25%.

Timeframe for the achievement of benefits It is expected that the agents will spread naturally from the initial points of establishment and colonise Californian thistle throughout the region. Without assistance, the time required for this spread to be achieved and for agents to reach damaging levels throughout the region is estimated at 10–20 years. However, if additional founding populations were established widely across the region it is estimated that the time required to achieve damaging population would fall rapidly. If the agents establish, it is certain that field days would be organised at which farmers could harvest agents for establishment on their own farms, further reducing the time lag between establishment and the realisation of benefits. It is therefore estimated that maximum annual benefits to farmers in the Otago/Southland will be achieved 10 years after the initial release.

If the agents are released in other parts of the country using the the same methods, then maximum annual benefits to farmers will be achieved 15 years after the initial release in Otago/Southland.

Costs and benefits The proposed control agents pose insignificant risk to non-target native plants. Thistles are rarely integral components of native ecosystems and the risks to native biodiversity and ecosystem processes are considered to be low.

The cost of completing this project is estimated to be $140,000 over the next three years. Moller et al. (2006) have provided a recent and reliable estimate of the combined cost of control of and production losses from Californian thistle on South Island sheep and beef cattle farms. This amounts to $8.3 million (or $2.9 million for Otago and Southland). Using these figures alone, there is large asymmetry between the actual costs of the project and the potential benefits to be achieved. Other estimates (table above) are either not recent, or are subject to major assumptions (Appendix B), but nevertheless indicate that potential benefits from improved Californian thistle are large, even in Otago and Southland. Even these underestimate the possible value of this project as the data presented here do not quantify most potential benefits (see above).

Given the low project cost, the low environmental risk, the low potential monetary risk and the massive potential monetary benefits. It is clear that the benefits of the proposal outweigh the risks and costs, without further cost– benefit analysis.

Section Eight – Satisfaction of the Section 36 Minimum Standards Satisfaction of the minimum standards in section 36 of the Act is a requirement for approval and will always be considered prior to the overall assessment and weighing of risks to, costs and benefits. Provide a statement in each subsection below on satisfaction of the minimum standards. Cross reference as appropriate (i.e. no need to repeat) to the detailed identification and assessments of risks set out in sections 6 and 7 above.

8.1 Displacement of native species State (with reasons) whether the new organism(s) is likely to cause any significant displacement of any native species within its natural habitat.

No significant displacement of any native species in its native habitat is envisaged either from the interaction of the two proposed agents with elements of the native flora or fauna (see Section 7.1), or from the expected reduction in thistle abundance.

8.2 Deterioration of natural habitats State (with reasons) whether the new organism(s) is likely to cause any significant deterioration of natural habitats.

There is no record of thistles forming a significant structural or ecological component of a native habitat in New Zealand. Tests indicate that these species will not attack native plants if released in New Zealand. Reduction of thistle abundance will therefore have no significant effect on natural habitats.

8.3 Adverse effects on human health and safety State (with reasons) whether the new organism(s) is likely to cause any significant adverse effects on human health and safety.

No significant effects on human health and safety are expected from the introduction of these species to New Zealand. They do not attack, sting, bite, or have an offensive odour, and are not known as vectors of human diseases. Many similar and related species already occur in New Zealand without raising such concerns.

8.4 Adverse effect to New Zealand’s inherent genetic diversity State (with reasons) whether the new organism is likely to cause any significant adverse effect to New Zealand‟s inherent genetic diversity.

There are no species of Ceratapion or Cassida currently resident in New Zealand. It is extremely unlikely that Ceratapion onopordi or Cassida rubiginosa could or would mate with a species of another genus, and therefore successful hybridisation is impossible. No such hybridisation is recorded in the scientific literature from Europe, where these species are native.

8.5 Causing disease, being parasitic, or becoming a vector for disease State (with reasons) whether the new organism(s) is likely to cause disease, be parasitic, or become a vector for human, animal, or plant disease. If, however, the purpose of the importation or release is to import or release an organism to cause disease, be a parasite, or a vector for disease all you need to do is state that.

Neither of these species is parasitic, or capable of directly transmitting human or animal diseases.

The capacity for Ceratapion onopordi to transmit spores of the Californian thistle rust fungus Puccinia punctiformis from stem to stem of Californian thistle has been acknowledged. It is therefore possible that these species could assist

Application for approval to import for release or release ER-AF-NOR-1-2 09/05 from containment any new organism including a genetically modified organism but excluding conditional Page 45 release and rapid assessment, under section 34 of the Hazardous Substances and New Organisms Act 1999 the spread of other plant diseases (e.g. rust fungi) by transporting spores. However, both feed on thistles only, and are very unlikely to transmit disease to non-thistle species via mouth parts. Ceratapion adults are small and Cassida adults are large but smooth. Neither species appear to have any biological, morphological or ecological features that would enhance the risk of passive transport, and are no more likely to transfer disease than any other insect. These species will make up only a small proportion of the total invertebrate fauna presumably capable of carrying spores, and any marginal increase in effects would be insignificant.

Section Nine – Additional Information

9.1 Do any of the organism(s) need approvals under any other New Zealand legislation or are affected by international obligations? For example, indicate whether the organism may be subject to other New Zealand legislation, e.g. the Biosecurity Act 1993, or Animal Welfare Act 1999; or if the organism(s) are listed in CITES, then approval is required from both the importing and exporting countries.

Neither species is listed by CITES (http://www.cites.org/eng/resources/species.html). Both are common, widespread species in Europe. There are no known regulatory requirements for export from Europe. Permits to import these species into New Zealand will be required under the Biosecurity Act 1993.

9.2 Have any of the new organism(s) in this application previously been considered for any form of approval in New Zealand or elsewhere? For example, has the organism(s) been previously considered for import (e.g. under the Plants Act)?

Neither Cassida rubiginosa nor Ceratapion onopordi has been intentionally released for the biological control of thistles anywhere in the world.

An application (NOC02004) to import Ceratapion onopordi (as Apion onopordi) into containment for further evaluation was approved by ERMA. The approval code was NOC002279. Import Permit no. 2006028253 was issued by MAF, and a population is presently held in the Landcare Research containment facility at Lincoln.

Cassida rubiginosa has not been the subject of any previous applications.

9.3 Is there any additional information that you consider relevant to this application that has not already been included?

9.4 Provide a glossary of scientific and technical terms used in the application

Abdomen hind section of an insect‟s body Assessment measuring impacts of biocontrol agents Asteraceous belonging to the daisy family Biennial taking two years to develop and seed Biological control the use of one living organism to control another Bolting growing a central stem from the centre of a rosette Clade Natural grouping of species, genera or tribes based on common ancestry (phylogeny)

Co-evolution where two organisms have evolved together and each influence the evolution of the other Coleoptera beetles Crepuscular active at dawn and dusk Crown the centre of the rosette, the top of the root Development test how far can a larva develop before dying when fed on a test plant – as opposed to a „starvation test‟ Diptera flies Dorsal relating to the upper surface

Elytra wing cases formed from one pair of wings Endemic naturally occurring in a country and nowhere else Epizootic parasitic on animals (including insects) from the outside or on the surface et al. and others Family a natural grouping of tribes or subfamilies Frass faecal material Food web the feeding relationship between various species living together in one area, the interaction of plants, herbivores, their parasitoids, predators and diseases. Frugivorous fruit-eating Gall a complex and characteristic structure grown by the plant in response to feeding by an eriophyid mite, and within which the mite lives Genus (plural Genera) a natural grouping of related species Herbaceous herb-like, non-woody, soft Herbivorous feeding on plants Host range the range of plants that a biocontrol agent can feed and reproduce on Host specificity testing testing to find out the host range of a potential biocontrol agent Hymenoptera wasps Indigenous native, but may occur elsewhere. Instar growth stage of an insect (in between moults), e.g. newly hatched = first instar Lag phase the period before a species becomes invasive in a new environment Larva(e) juvenile stage(s) of an moth Lateral relating to the side Maxillary palps appendages near the mouth Monophagous feeding on only one species of plant Monospecific only feeding on one plant species Multivoltine having multiple generations per year

Oligophagous feeding on just a few closely related species of plants Ootheca egg case Oviposit lay eggs Oviposition egg-laying Parasitoid an insect that feeds and develops on or within another living host insect. It completes its own development on a single host, which it kills in the process Pers. comm. personal communication Phenology the life history of the insect and how it reacts to environmental variables such as temperature Photosynthesis process by which plants harness sunlight to build carbohydrates Phylogeny The relationship between species based on common ancestry rather than physical similarity, and determined by genetic analysis Phytophagous plant feeding Pinnacula small bumps in the skin of a caterpillar.

Prothoracic plate shield-like plate immediately behind the caterpillar‟s head Pupa(e) stage(s) of insect development between larva and adult Rosette juvenile plant, before the central tall flower stalk begins to grow Species a morphologically, behaviourally, or ecologically distinct group of individuals that can only breed successfully with its own kind Starvation test does a newly-hatched larva die rather than feed on a test plant? – as opposed to a „development test‟ Sympatry overlapping of natural ranges of two populations Taxa in this context, the total of species and variants of species Tribe a natural grouping of genera

9.5 List of appendices. List any appendices included with this application. Any information that is commercially sensitive or additional material included with the application (such as details of consultations, referenced articles) should be contained in appendices. The main application should refer to the relevant appendices but be able to be read as a stand-alone document along with the cover page of the book.

Appendix A. Consultation with the community prior to application The scope of consultation Consultation with Māori Responses from Iwi, Hapū and other Māori organisations Consultation with other organisations Personal communications and submissions The relationship between New Zealand daisy species The „request for consultation‟ and the „information pack‟

Appendix B The biology, weed status and economics of Californian thistle in New Zealand The weed status of Californian thistle in New Zealand Current control methods The cost of Californian thistle control in Otago and Southland The cost of Californian thistle control elsewhere ein New Zealand Effects on primary production in Otago and Southland Effects on primary production elsewhere in New Zealand The relationship between control costs and gains in productivity Other sources of losses

Appendix C Host specificity of Cassida rubiginosa Müll. (Coleoptera: Chrysomelidae) and Apion onopordi Kirby (Coleoptera: Apionidae), potential control agents for Californian thistle (Cirsium arvense) in New Zealand

9.6 References. Please include a list of the references cited in and supplied with this application form. Originals of the references must be supplied in full. Where the reference supplied is an extract from a book only the specific pages quoted must be supplied.

Ang BN, Kok LT, Holtzman GI, Wolf DD 1995. Canada thistle (Cirsium arvense (L.) Scop.) response to density of Cassida rubiginosa Müller (Coleoptera: Chrysomelidae) and plant competition. Biological Control 5: 31–28.

Bourdôt G, Kelly D 1986. Density and cover estimates of some non-palatable herbaceous pasture weeds. Proceedings of the 39th New Zealand Plant Protection Conference: 183–186.

Briese DT 2003. The centrifugal phylogenetic method used to select plants for host-specificity testing of weed biological control agents: Can and should it be modernised? In:

Briese DT, Walker A 2002. A new perspective on the selection of tests plants for evaluating the host-specificity of weed biological control agents: the case of Deuterocampta quadrijuga, a potential insect control agent of Heliotropium amplexicaule. Biological Control 25: 273–287.

*Briese DT, Heard TA, McFadyen RE, Sheppard AW, Spafford Jacob H. 2003. The selection, testing and evaluation of weed biological control agents: is there still room for improvement. CRC for Australian Weed Management Technical Series 7. Pp. 23–33.

*CABI Crop Compendium 2004: http://www.cabi.org/compendia/cpc/index.htm

Crop and Food Research 2001. Globe artichokes – an ancient crop offering new opportunities. Broad sheet 103.

Eerens JP, Seedfelt SS, Carry G, Armstrong ML 2002. Controlling Californian thistle (Cirsium arvense) through pasture management. New Zealand Plant Protection 55: 111–115.

*ERMANZ 2004. Evaluation and Review Report NOC02004.

Freese G. 1995. The insect complexes associated with the stems of seven thistle species. Entomologia Generalis 19: 191–207.

Friedli J, Bacher S 2001. Direct and indirect effects of a shoot-base boring weevil and plant competition on the performance of creeping thistle, Cirsium arvense. Biological Control 22: 219–226.

Funk VA, Bayer AJ, Keeley S, Chan R, Watson L, Gemeinholzer B, Schilling E, Panero JL, Baldwin BG, Garcia-Jacas N, Susanna A, Jansen RK 2005. Everywhere but Antarctica: using a supertree to understand the diversity and distribution of the Compositae. Biol. Skr 55: 343–374.

Gassman A 2005. Classification of thistle species and potential for biological control of Cirsium arvense in New Zealand. Special report on behalf of Landcare Research New Zealand Ltd. An unpublished report to Landcare Research by CABI Bioscience, 24 p.

Gassman A, Tosevski I, Petanovic R, Magud B, Häfliger P, Chevillat V, Rheinhold T 2005. Biological control of Canada thistle (Cirsium arvense) – Preliminary Annual Report 2005. An unpublished report to Landcare Research by CABI Bioscience. 9 p.

Gourlay AH 2003. Host specificity of Apion onopordi Kirby (Coleoptera: Apionidae), a potential control agent for Californian thistle (Cirsium arvense) in New Zealand. Landcare Research Report LC0304/003, unpublished. 12 p.

*Hackwell K, Bertram G 1999. Pests and weeds: A blueprint for action. New Zealand Conservation Authority.

Harris, P. 2005. Cassida rubiginosa Müller. Defoliating beetle. http://res2.agr.ca/lethbridge/weedbio/agents/acasrub_e.htm

*Julien MH, Griffiths MW 1998. Biological control of weeds: A world catalogue of agents and their target weeds, 4th edition. Wallingford, UK, CAB International.

*Kok LT, McEvoy PJ, Mays WT 2000. Successful establishment of exotic agents for classical biological control of invasive weeds in Virginia In: Spencer NR ed. Proceedings of the Xth International Symposium on Biological Control of Weeds. 4–14 July 1999, Montana State University, Bozeman, Montana, USA. Pp. 59–65.

*Kuschel G. 2003. Nemonychidae, Belidae, Brentidae (Insecta: Coleoptera: Curculionoidea). Fauna of New Zealand 45: 1–100.

Labes J 2000. Californian thistle. A report to AgMARDT on a survey conducted for the Californian Thistle Action Group. 31 p.

McClay AS 2002. Caanada thistle In: Van Driesche R et al. Biological control of invasive plants in the Eastern United States. USDA Forest Service publication, FHTET-2002-04, 413p.

*MAF Statistics. http://www.maf.govt.nz/statistics/primary-industries/index.htm

Martin RJ, Lauren DR, Smith WA, Jensen DJ, Deo B, Douglas JA 2006. Factors influencing slymarin content and composition in variegated thistle (Silybum marianum). New Zealand Journal of Crop and Horticultural Science.34: 239–245.

Mitchell RB, Abernethy RJ 1993. Integrated management of Californian thistle in pasture. Proceedings of the 46th New Zealand Weed and Pest Control Conference: 278–281.

* Moller H, Hill RL, Lucock D 2006. A survey of weed management on South Island sheep and beef farms. Extracts and summaries from a draft report to the Sustainable Farming Fund (see Appendix B)

Popay AI, Rahman A, James TK 2002. Future changes in New Zealand‟s hill country pasture weeds. New Zealand Plant Protection 55: 99–105.

Rahman A, Popay AI 2001. Review of emerging weed problems in hill country pastures. A report commissioned by MAF Policy. http://www.maf.govt.nz/mafnet/rural-nz/sustainable-resource-use/land-management/emerging- weeds/

Reitz SR, Trumble JT 2002. Competitive displacement among insects and arachnids. Annual Review of Entomology 47: 435–465.

Shea K, Kelly D 2004. Modelling for management of invasive species: musk thistle (Carduus nutans) in New Zealand. Weed Technology 18: 1338–1341.

Teulon DAJ, Cameron PJ, Bourdot GW, Curtin D 2005. Plant protection in organic arable and vegetable crops. Lincoln, Crop and Food Research.

Van Toor RF, Stuck RJ 1993. Farmers‟ evaluation of weed and invertebrate pest problems in pasture. Proceedings of the 46th New Zealand Plant Protection Conference: 200–205.

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

*Zwölfer H 1965. Preliminary list of phytophagous insects attacking wild Cynareae (Compositae) in Europe. Technical Bulletin 6, Commonwealth Institute of Biological Control, Delemont, Switzerland. Pp. 81–154.

Zwölfer H, Eichhorn O 1966. The host ranges of Cassida spp. (Col: Chrysomelidae) attacking Cynareae (Compositae) in Europe. Journal of Applied Entomology 58: 384–397.

* = reference not provided because it is unpublished, too large, a book, a website, or an ERMA publication.

Section Ten – Application Summary Summarise the application in clear, simple language that can be understood by the general public. Include a description of the organism(s) to be released, and any risks, costs and benefits associated with their release. Any consultation that was undertaken should be noted. This summary will be used to provide information for those people and agencies who will be notified of the application (e.g. Ministry of Agriculture and Forestry, Department of Conservation, Ministry for the Environment etc) and for members of the public who request information. Do not include any commercially sensitive information in this summary – this should be attached as a separate Appendix and clearly marked “confidential”.

The Californian Thistle Action Group wishes to introduce two insects to help control Californian thistle; the green thistle beetle and the thistle stem weevil. The application to introduce these control agents incorporates input from wide consultation with Māori, companies, local government, and conservation and producer organisations.

Californian thistle is not a great problem in native habitats but is probably New Zealand‟s most costly weed in pastures and crops. Most thistles come and go, but Californian thistle is perennial, forming long-lasting patches that creep outwards year by year. Buds on the deep, massive root system push up and produce a rosette in spring, and then grow up to form dense stands of prickly stems that shade out crop and pasture plants from late spring. Sheep refuse to graze around the stems, reducing the amount of pasture available to grazing stock, reducing stock numbers, and so limiting returns for farmers. If you knock the thistle stems down two or three times a year by mowing, the root system eventually becomes exhausted, and patches decline (Eerens et al. 2002). Selectively applying herbicide using a „weedwipers‟ does the same thing. However, this only works where you can safely drive a tractor or vehicle, and isn‟t possible on steeper country. There farmers often have to resort to aerial application if they want any production from infested land.

It is hard to estimate the losses to New Zealand agriculture from the costs of control and the loss in production caused by Californian thistle. We know that for South Island sheep beef farms alone the cost is $8.3 million. The thistle affects all types of farm nationwide, and so for the whole of New Zealand pastoral agriculture the cost is likely to be hundreds of millions. It severely hampers crop production as well, and biological control would bring relief to the horticulture industry.

Grazing by biological control agents could achieve Californian thistle control just as repeated mowing does, but would not be restricted to flat land, and would be achieved without repeated treatments. The green thistle beetle introduced itself to North America many years ago, and often causes decline in the density of thistle stems. The stem weevil has not been used as a biological control agent before, but is expected to assist the beetle, as well as spreading a rust fungus of Californian thistle from stem to stem. Both insects attack other thistles, and are expected to reduce the effects of nodding thistle and Scotch thistle in pastures as well.

The insects are well known in Europe and have never been recorded as pests of any plants there. Although both insects feed on more than one plant, tests indicate that they pose no significant risk to native plants. The most closely related New Zealand plants are puha species which are highly prized by Māori. Neither species attacks puha. Several other desirable plants are closely related to thistles. Globe artichoke and safflower were slightly damaged by the insects in experiments. The insects have never been recorded as pests of these crops in their native Europe, so the risk to these crop plants in New Zealand is considered to be low. Variegated thistle has been suggested as a commercial crop here, and this would certainly be attacked by both species. It is likely that ornamental cornflowers growing in domestic gardens will also receive some damage from both agents.

The application identifies and assesses the risks and benefits of this proposal. Overall, the potential benefits to agriculture that would arise from the successful biological control of Californian thistle are very large, and the prospects for achieving a level of control using these insects are good. The potential benefits of this project greatly

Checklist

Please check and complete the following before submitting your application:

All sections completed Yes Appendices enclosed Yes/ NA* Confidential information identified and enclosed separately Yes/NA Copies of additional references attached Yes/NA Cheque for initial fee enclosed (incl. GST) Yes/No If “yes”, state amount: $………. Direct credit made to ERMA bank account: Yes/No If „yes” give date of DC …/…/… and amount: $………. Application signed and dated Yes Electronic copy of application e-mailed to ERMA New Zealand Yes

*NA – not applicable

† The cost of the application (our fee) can be found on our web site under new organism applications.

Signed: Date: