Environment Protection and Biodiversity Conservation Act 1999

Referral of proposed action

Project title: Importation, rearing and release of Tachardiaephagus somervillei (Hymenoptera: Encyrtidae) as a biological control agent for the yellow lac scale Tachardina aurantiaca (Hemiptera: Kerriidae) on Christmas Island, Indian Ocean

1 Summary of proposed action NOTE: You must also attach a map/plan(s) showing the location and approximate boundaries of the area in which the project is to occur. Maps in A4 size are preferred. You must also attach a map(s)/plan(s) showing the location and boundaries of the project area in respect to any features identified in 3.1 & 3.2, as well as the extent of any freehold, leasehold or other tenure identified in 3.3(i).

1.1 Short description Use 2 or 3 sentences to uniquely identify the proposed action and its location. Supercolonies of the invasive yellow crazy ant Anoplolepis gracilipes (YCA) are a major and on- going threat to biodiversity values on Christmas Island, especially to red land crabs Gecarcoidea natalis and robber crabs Birgus latro. To date, the management of YCA supercolonies has depended on surveillance, monitoring and control using toxic bait (mostly fipronil), particularly through aerial baiting programs in 2002, 2009 and 2012. While this program has been very effective in suppressing YCA supercolonies and there are encouraging signs of recovery in many treated areas, new supercolonies continue to form. There is widespread concern for the sustainability of this program in terms of its expense, non-target impacts, and the resources it diverts from other conservation programs. Recent research conducted by La Trobe University, funded by the Director of National Parks and endorsed by the Christmas Island Crazy Ant Scientific Advisory Panel, indicates that long-term, sustainable suppression of YCA supercolonies could be achieved through the introduction of a host-specific biological control agent that would indirectly affect YCA by reducing carbohydrate supply provided by scale insects, a key resource implicated in supercolony dynamics. The yellow lac scale Tarchardina aurantiaca (Hemiptera, Kerriidae; hereafter Tachardina) is likely to be the single biggest contributor to the honeydew economy of YCA supercolonies across the island. This species, like all other scale insect species on the island, is not native to Christmas Island. The proposed action is to import, rear and release on Christmas Island a key natural enemy of Tachardina, the parasitoid microhymenopteran wasp Tachardiaephagus somervillei (Hymenoptera, Encyrtidae; hereafter Tachardiaephagus) as a biological control agent for this scale insect. The expectation is that Tachardiaephagus should indirectly suppress supercolony formation by YCA by controlling its key scale insect mutualist.

1.2 Latitude and longitude Latitude Longitude Latitude and longitude details location point degrees minutes seconds degrees minutes seconds are used to accurately map the boundary of the proposed action. If these coordinates are NE extreme -10 24 40 105 42 14 inaccurate or insufficient it may NW extreme -10 26 14 105 32 59 delay the processing of your SW extreme -10 30 54 105 32 1.6 referral. S extreme -10 34 41 105 39 53 E extreme -10 28 4 105 42 49

Page 1 of 44 Environment Protection and Biodiversity Conservation Act 1999

1.3 Locality and property description Provide a brief physical description of the property on which the proposed action will take place and the project location (eg. proximity to major towns, or for off-shore projects, shortest distance to mainland).

The project will be carried out on Christmas Island, an Australian External Territory located in the northeastern Indian Ocean at 10° 25’S and 105° 40’E (Fig. 1a). The island is approximately 2600 km northwest of Perth and 360 km south of the western end of Java, Indonesia. The island covers 135 km2, of which approximately 85 km2 (64%) is National Park (Fig. 1b). The remaining area is unallocated Commonwealth Crown Land (18%); mine lease (15%) (leased by Phosphate Resources Limited from the Commonwealth) and committed land, including private dwellings (3%). The Park contains two Ramsar wetlands at The Dales and Hosnie’s Spring (Fig. 1b).

Christmas Island rises from the sea in a series of cliffs and terraces to a maximum elevation of 361 meters. The terraces are separated by slopes covered by loose and jagged limestone boulders, or by sheer cliffs (Fig. 1c). The interior of the island is a slightly undulating plateau, from 160-361 metres above sea level (Fig. 1d). About 75% of the island is covered by vegetation, mostly rain forest.

Figure 1. (a) Geographic location of Christmas Island in the north eastern Indian Ocean, (b) Christmas Island showing the National Park (green), unallocated crown land (yellow), Ramsar sites (Blue hatching) and phosphate mining leases (green stippling). The National Park and Unallocated Crown Land together show the extent of remaining rainforest on Christmas Island, (c) Christmas Island rises in a series of cliff and terraces to a maximum elevation of 361 m (d) the interior of the Island is a slightly undulating plateau, from 160-361 m above sea level.

(b) (a)

(c) (d)

Page 2 of 44 Environment Protection and Biodiversity Conservation Act 1999

1.4 Size of the development YCA supercolonies have formed patchily in rainforest across the footprint or work area (hectares) entire Island (Fig. 2). The intention is for the biological control agent to both control existing supercolonies and suppress the formation of new ones, by eventually colonising and establishing self-sustaining populations across the entire forested area from an initially small number of release points. Thus, the proposed action is forecast to occur over the entire forested area of the Christmas Island National Park and adjacent unallocated Crown Land, totalling c. 10,000 ha.

Figure 2. Composite map showing YCA Supercolonies treated with toxic bait between 2000-2012. This map effectively shows the extent of supercolony formation over the last twelve years.

1.5 Street address of the site Christmas Island National Park, PO Box 867, Drumsite, Christmas

Island, Indian Ocean, 6798. Unallocated Crown Land, managed by the Commonwealth Department of Department of Regional Australia, Local Government and Sports (DRALGAS).

1.6 Lot description Describe the lot numbers and title description, if known. N/A

1.7 Local Government Area and Council contact (if known) If the project is subject to local government planning approval, provide the name of the relevant council contact officer. Although local government approval is not required for the proposed action, consultation will still occur. The relevant local government is the Shire of Christmas Island – Western Australia Local Government Association. CEO Kelvin Mathews.

Page 3 of 44 Environment Protection and Biodiversity Conservation Act 1999

1.8 Time frame Specify the time frame in which the action will be taken including the estimated start date of construction/operation.

Foreign exploration in Southeast Asia has already located Tachardina at sites across 1900 km in Malaysia, within its native range in Southeast Asia. Research on the biology of its main parasitoid Tachardiaephagus is already underway in Southeast Asia, principally at the Forest Research Institute of Malaysia in Kuala Lumpur and at Sarawak Forestry in Semenggoh, Sarawak (see O’Dowd et al. 2012, and Section 2.1 below). A preliminary draft for host specificity testing has been produced (see Attachment 1) and host specificity testing could be completed in Malaysia by June 2013. Coincident with these activities, rearing facilities for Tachardiaephagus will be constructed on Christmas Island (see Attachment 2). Tachardiaephagus will be imported to Christmas Island when both tasks are completed, and a population will be established under controlled laboratory conditions in the rearing facility. The first field releases are planned for mid- 2014, and there will be multiple releases thereafter through 2015 and 2016.

1.9 Alternatives to proposed No action Were any feasible alternatives to taking the proposed action (including not taking the action) considered but are not proposed? Yes, you must also complete section 2.2

1.10 Alternative time frames etc Does the proposed action No include alternative time frames, locations or activities? Yes, you must also complete Section 2.3. For each alternative, location, time frame, or activity identified, you must also complete details in Sections 1.2-1.9, 2.4-2.7 and 3.3 (where relevant). 1.11 State assessment No Is the action subject to a state or territory environmental Yes, you must also complete Section 2.5 impact assessment? NOTE: As described in Section 2.4, the regulatory framework under which the introduction of a biological control agent to the external Australian territory of Christmas Island would be considered cannot yet be clearly identified. Therefore, the requirement for a formal state or territory environmental impact assessment is also unclear. The matters of greatest relevance in the assessment of this Referral are the host specificity of the proposed biological control agent, and the likelihood of off-target impacts on other scale insects or species of concern. These issues are addressed in Section 2.5 in lieu of a formal environmental impact assessment.

1.12 Component of larger action No Is the proposed action a component of a larger action? Yes, you must also complete Section 2.7

1.13 Related actions/proposals Is the proposed action related to other actions or proposals in No the region (if known)?

Yes, provide details:

Page 4 of 44 Environment Protection and Biodiversity Conservation Act 1999

1.14 Australian Government No funding Has the person proposing to Yes, provide details: The research supporting the proposed action take the action received any is being funded by DSEWPaC though the Director of National Parks Australian Government grant to La Trobe University. One postdoctoral fellow has just finished a funding to undertake this three-year project establishing the importance of liquid carbohydrate project? resources provided by scale insects for YCA supercolony formation (November 2009 – November 2012), while a second postdoctoral fellow has been working on research and survey for prospective biological control agents of Tachardina (June 2010 – June 2013). This postdoctoral researcher has spent considerable time in Southeast Asia. It is intended that DSEWPaC continue to fund La Trobe University through the host-specificity testing phase in Kuala Lumpur, and on to the importation, rearing and release phases on Christmas Island (July 2013 – July 2016).

1.15 Great Barrier Reef Marine Park No Is the proposed action inside the Great Barrier Reef Marine

Park? Yes, you must also complete Section 3.1 (h), 3.2 (e)

2 Detailed description of proposed action NOTE: It is important that the description is complete and includes all components and activities associated with the action. If certain related components are not intended to be included within the scope of the referral, this should be clearly explained in section 2.7.

2.1 Description of proposed action This should be a detailed description outlining all activities and aspects of the proposed action and should reference figures and/or attachments, as appropriate.

Background – Biology and Impacts. Anoplolepis gracilipes (the yellow crazy ant, YCA) is a ‘tramp’ species that has become invasive throughout the tropics (Wetterer 2005). The YCA is listed by the IUCN as one of the world’s 100 worst invasive species (Lowe et al. 2000), and was accidentally introduced to Christmas Island between 1915 and 1934 (O’Dowd et al. 1999). Because of its negative impacts on species, interactions and ecosystem processes, YCA is recognised as the most significant and pervasive threatening process affecting biodiversity on Christmas Island, reflected by the listing of the Loss of biodiversity and ecosystem integrity following invasion by the Yellow Crazy Ant on Christmas Island, Indian Ocean as a Key Threatening Process under the EPBC Act 1999, and as identified in Threat abatement plan to reduce the impacts of tramp ants on biodiversity in Australia and its territories (Commonwealth of Australia 2006). The control of this ant features prominently in many Recovery Plans for EPBC-Listed species on Christmas Island, and has been the focus of natural resource management activities on the island for more than a decade at the cost of millions of dollars. The attribute that makes the YCA so problematic on Christmas Island and elsewhere is its capacity to form high-density, expansive ‘supercolonies’. Many tramp ant species form supercolonies, often defined using a combination of criteria including genetic relatedness,intraspecific behavioural interactions, and ant abundance. For example, the invasive argentine ant Linepithema humile is genetically uniform across most of its European range and is regarded as a single supercolony (Giraud et al. 2002), while intercontinental behavioural assays and analyses of cuticular hydrocarbons suggest this species may form a single supercolony spanning Australia, Europe, North

Page 5 of 44 Environment Protection and Biodiversity Conservation Act 1999

America and Japan (Sunamura et al. 2009; Suhr et al. 2010). Although two distinct genotypes of YCA occur on Christmas Island, these co-occur at very small spatial scales (Thomas et al. 2010) and behavioural assays pairing individual ants from opposite ends of the island suggest that the population on Christmas Island behaves as a single supercolony (Abbott 2005). Nevertheless, YCA supercolonies on Christmas Island have always been defined in terms of very high ant densities. YCA occurs at many locations across the island in very low abundance with no obvious impact on biodiversity, but in 1989, and then again in late 1997, YCA was discovered in several locations at extremely high densities sufficient to extirpate local populations of the abundant red land crab Gecarcoidea natalis (O’Dowd et al. 1999, 2003). The red crab is a keystone species in rainforest on the island that controls patterns seedling recruitment and rates of litter breakdown and nutrient cycling (Green et al. 1997, 2008). Since then, the density at which YCA kills land crabs leading to understorey transformation has become the operational definition of ‘supercolony’ on Christmas Island. On the ground, the density of YCA in supercolonies can be extraordinary – up to several thousand per square meter (Abbott 2005). A systematic, island-wide survey in 2001 found multiple supercolonies ranging in area from tens to hundreds of hectares, totalling c. 2500 ha, or 25% of rainforest on the island (Green et al. 2004; Green and O’Dowd 2009). Since 2001, supercolonies have continued to form and/or reform and upwards of 5000 ha of rainforest have been treated with toxic bait to 2012. The key trait that has allowed YCA to form high-density supercolonies is its ability to form mutualistic associations with honeydew-producing hemipterans, principally scale insects (O’Dowd et al. 2003; Abbott and Green 2007). The scale insects suck sap from trees and secrete carbohydrate-rich honeydew on which the ants feed. The ants provide sanitation services for the scale insects, removing honeydew that might otherwise build up and kill them either through asphyxiation or the growth of sooty moulds, but the ants may also provide limited protection for the scale insects from generalist natural enemies. Supercolony-level densities of YCA and outbreak-densities of several species of scale insects invariably co-occur, and in supercolonies high densities of ants can typically be seen ascending the boles of most trees to visit scale insects in the canopy. The gasters of descending ants are swollen with carbohydrate-rich honeydew that they will take back to the nest to be shared with non-forager conspecifics. Site-scale, manipulative experiments on Christmas Island have demonstrated a bi-directional, causal link between co-occuring high densities and ants and scale – the exclusion of ants using toxic bait leads to a dramatic decline in scale abundance (Abbott and Green 2007), while the prevention of access by ants to scale insects using tree bands leads to a dramatic decline in ant density (O’Dowd et al. 2012). The tree banding experiment especially is important as “proof of concept” that removing a key food resource for ants could lead to supercolony suppression. On Christmas Island YCA interacts with several species of honeydew-secreting scale insects including Tachardina aurantiaca, the principal target of the biocontrol program (see below). The mutualism between YCA and scale insects has manifold negative impacts on species abundances, interactions among species, and forest structure. At the core of these impacts is the devastating effect of YCA on land crabs, especially red crabs. YCA sprays formic acid as a weapon both to subdue prey and in self-defence, and although the amount sprayed by individual ants is tiny, at supercolony densities the overall effect is overwhelming. YCA supercolonies reduce formerly high densities of red crabs (averaging c. 0.5 – 1.0 crabs m-2) to nil, deregulating seedling recruitment and litter dynamics and resulting in a thick, diverse understorey of seedlings and saplings with an almost permanent layer of leaf litter (O’Dowd et al. 2003). In forest dominated by red crabs, the understorey is sparse and dominated by a few crab-resistant species, and the forest floor is almost devoid of litter for much of the year (Green et al. 1997, 2008). These impacts are widespread. Based on the spatial extent of supercolony formation over the last 12 years, it is likely that YCA has extirpated at least 20 million red land crabs, or about 30% of the total population in areas where they have formed supercolonies.

Page 6 of 44 Environment Protection and Biodiversity Conservation Act 1999

YCA has also caused declines in the density of red land crabs at sites where high-density supercolonies have never formed. About half of the red crab population migrates to the coast each year to complete breeding activities, and many YCA supercolonies have formed across the crabs’ traditional migration routes. Thus, significant numbers of migrating red crabs have been killed en route to the coast over many years, never to return to the former home ranges. As a result, some areas of rainforest are practically devoid of red crabs even though YCA supercolonies have never occurred there, and the same processes of understorey transformation are in train there too. It is hard to gauge the severity and extent of this “ghosting” effect because pre-YCA invasion data on red crab densities are sparse, but is likely to be significant; Green et al. (2011) estimated that around 25% of rainforest may have been ghosted at some time in the last decade. The direct and indirect (ghosting) impacts of YCA supercolony formation have been so widespread since the late 1990s that just 28% of rainforest could still be considered as “intact” (no YCA supercolony formation, and unaffected by ghosting) by 2007 (Green et al. 2011). In YCA supercolonies, scale insects themselves can have large negative impacts on their host plants. Especially vulnerable is the Tahitian Chestnut Inocarpus fagifer, a widespread canopy tree that hosts very high densities of Tachardina on its outer twigs and leaves. In supercolonies, seedlings, saplings and small trees all suffer extremely high mortality, and the canopies of large trees are much reduced through the dieback of fine twigs and branches (Green et al. 2001, O’Dowd et al. 2003, P. Green and D. O’Dowd, unpublished results). There is also evidence that fruit production is reduced in supercolonies. Excess honeydew that YCA does not harvest settles on leaves of all plant species and is colonised by sooty moulds, which probably interferes with photosynthesis and growth. YCA may affect many of the island’s bird species through direct interference and through altered resource availability and habitat structure (Davis et al. 2008, Davis et al. 2009). The Christmas Island Emerald Dove Chalcophaps indica natalis is 9-14 times less abundant in ant-invaded forest, and because it forages on the forest floor, is probably vulnerable to direct predation by YCA. The nesting success and density of juvenile Christmas Island Thrushes Turdus poliocephalus erythropleurus is lower in supercolonies, where they also show altered foraging and reproductive behaviours. Furthermore, these birds alter their choice of tree species in which to build nests, with lower frequency on tree species that typically experience high densities of scale insects and ants. The density of foraging Christmas Island white-eye Zosterops natalis is higher in supercolonies, perhaps because scale insects (as prey) are more common there. It is possible that impacts of YCA on thrushes and white-eyes affect frugivory and seed dispersal on the island; assays with both real and model fruit showed handling rates to be more than two-fold lower in YCA supercolonies, and manipulative experiments showed this to be a direct consequence of the presence of ants. There is no evidence that YCA supercolony formation significantly affects the density of nesting success of Abbott’s Booby Papasula abbotti (P. Green, unpublished data), while the impact of YCA on other seabirds and on other land birds such as the goshawk and owl are unknown. Several endemic vertebrate species, including the pipistrelle bat Pipistrellus murrayi and endemic reptiles have experienced precipitous declines over recent years, but the causes of the declines are opaque. In the case of the pipistrelle it is possible that YCA supercolony formation has contributed to the decline, either directly through predation of bats at roost sites or indirectly by eliminating red crabs and facilitating the expansion of predators such as giant centipedes Scolopendra subspinipes, wolf snakes Lycodon aulicus, cats and rats (Schulz and Lumsden 2004, Lumsden et al. 2007, Beeton et al. 2010). However, the decline of the pipistrelle was well in train before the rise of YCA supercolonies in the late 1990s. The endemic reptiles were similarly in decline long before YCA supercolonies became common, and the role of YCA in their decline is also uncertain (Smith et al. 2012).

Page 7 of 44 Environment Protection and Biodiversity Conservation Act 1999

In addition to impacts on species of concern, supercolony formation by YCA has also led to an altered web of species interactions that facilitates the entry and spread of other invasive species. The best example of this is the entry and spread of the giant African landsnail Achatina fulica (GALS) in rainforest on the island. GALS was introduced to the island in the 1940s, and despite being a notoriously invasive species (Lowe et al. 2000), its distribution was for many decades limited to settled areas, abandoned mining fields and roadsides. Experiments showed that predation by red crabs excluded this invader from establishing in rainforest (Lake and O’Dowd 1991), but the extirpation of red crabs in YCA supercolonies, coupled with the ability of GALS to coexist alongside YCA in supercolonies, has permitted this snail to establish high densities in rainforest at many locations across the island (Green et al. 2011). The facilitation of GALS by YCA could be due to the creation of enemy-free space, augmented understorey resources, or both. The rise of YCA supercolonies and the extirpation of red crabs have also affected the invasion dynamics of other non-native organisms. These effects encompass both inhibition and facilitation for a range of non-native ant and snail species (O’Dowd and Green 2010; P. Green and L. O’Loughlin, unpublished results), while invasion by several weeds including Capsicum frutescens, Carica papaya, Cordia curassavica, and Muntingia calabura appears to be facilitated in areas affected by YCA supercolony formation (P. Green and D. O’Dowd, personal observations). Background – Management. Given all the above, supercolony formation by YCA is considered a major and on-going threat to biodiversity values on Christmas Island. To date the management of YCA has depended on surveillance, monitoring, and control using toxic bait (Green and O’Dowd 2009, Boland et al. 2011). New supercolonies continue to form, and there is concern for the sustainability of this program in terms of its expense, non-target impacts, and the resources it diverts from other conservation programs. Further, this program can only ever be reactive, and it has not been able to find an effective solution to the difficult issue of the management of incipient supercolonies at an islandwide spatial scale (see section 2.2, below). There is widespread agreement that the development of a more cost-effective, sustainable alternative to the use of toxic bait is needed to manage the YCA invasion on Christmas Island. For the last three years, La Trobe University researchers have been investigating the potential for supercolony suppression by the biocontrol of scale insects. These researchers presented a summary of their findings to the Crazy Ant Scientific Advisory Panel (CASAP), and independent group of scientists who provide advice to the Director of National Parks and Parks Australia on matters pertaining to the management of the YCA invasion. CASAP advice was formally sought on the veracity of the research findings, and on the proposal to import, rear and release a biological control agent(s) as an option for controlling YCA supercolony formation and spread. At a meeting in December 2012, CASAP assessed the scientific merit of the research conducted by La Trobe University as well as the feasibility and risks of introducing a biological agent to Christmas Island. CASAP members were convinced that the option of biological control was a viable and feasible option for controlling YCA, the risks to the island’s biodiversity were very low, and the risk of doing nothing outweighed the risk posed by the importation and release of a biological control agent. CASAP advised the DNP to proceed with the implementation phase of the biological control research (see Attachment 3). Biological Control. Classical biological control works on the principle that in their area of origin, populations of native species are kept in check by their natural enemies (predators, parasites or pathogens). There is a large body of literature demonstrating that in many cases, species introduced outside of their native ranges become invasive because they have effectively left their enemies behind. This is known as the “Enemies Release Hypothesis”. The principle of classical biological control then is the converse: to re-establish population control over invasive species by first identifying and then importing a natural enemy – a biological control agent – from within the native

Page 8 of 44 Environment Protection and Biodiversity Conservation Act 1999

range of the target organism. In all instances, this now involves selection and testing of biological control agents to verify narrow host ranges that minimize the risk of non-target impacts in the area of introduction. Indirect Biocontrol for YCA on Christmas Island. Despite the diversity and significant ecological and economic impacts of invasive ants worldwide, they have proved to be an especially difficult target group for biological control. A program for the biocontrol of the Red Imported Fire Ant (Solenopsis invicta) using a parasitic fly and a protozoan disease as agents is currently under development in the southeastern United States, but no species of ants have yet been controlled in the field using biological control agents and principles. Instead, a novel solution for managing the YCA invasion is proposed; rather than targeting the ant itself, a key mutualist species that plays a significant role in sustaining YCA supercolonies at very high and ecologically damaging densities will be targeted instead. Long-term, sustainable suppression of YCA supercolonies could be achieved through the introduction of a biological control agent that would indirectly affect YCA by reducing carbohydrate supply provided by honeydew-producing scale insects, a key resource implicated in supercolony dynamics. The Proposed Action. While several species of honeydew-producing scale insects are common in YCA supercolonies, the yellow lac scale Tarchardina aurantiaca (Hempitera, Kerriidae) is almost certainly the single biggest contributor to the honeydew economy of YCA supercolonies across the island (O’Dowd et al. 2012). This referral is a proposal to import, rear and release on Christmas Island a key natural enemy of this scale insect, the parasitoid microhymenopteran wasp Tachardiaephagus somervillei (Hymenoptera: Encyrtidae - hereafter Tachardiaephagus) as a biological control agent for Tachardina. The expectation is that Tachardiaephagus will indirectly suppress existing YCA supercolonies and arrest formation of new supercolonies by controlling its key scale insect mutualist. The Proposed Biological Control Agent. Tachardiaephagus somervillei (Fig. 3) is the most promising agent for introduction and release on Christmas Island. It is one of several primary parasitoids that have been found to attack yellow lac scale during studies conducted within it native distribution in Malaysia (Table 1). This parasitoid

 attacks Tachardina across a 1900-km range in Peninsular Malaysia and Malaysian Borneo  is the most abundant natural enemy of Tachardina  has a short life cycle compared to its host  exhibits superparasitism (i.e., where multiple progeny emergence from a single host individual)  causes high rates of parasitism on Tachardina in the presence of tending ants, including the yellow crazy ant (Fig. 4)  can be reared under laboratory conditions.

Page 9 of 44 Environment Protection and Biodiversity Conservation Act 1999

Figure 3. Tachardiaephagus somervillei, a primary parasite of the yellow lac scale Tachardina aurantiaca in Malaysia and Singapore (drawing from Narayanan 1962. Pests of lac in India. Pp. 90-113 in Mukhopadhyay, B and Muthana, MS (eds) A Monograph on Lac. Indian Lac Res. Institut. Nancum, Ranchi, India). A Master’s student at Universiti Sains in Malaysia has also successfully reared T. somervillei under laboratory conditions.

1 mm

Figure 4. Parasitism of the yellow lac scale Tachardina aurantiaca in Malaysia. The image shows emergence holes of the parasitoid Tachardiaephagus sp. (Encyrtidae) in tests of an aggregate of old adult females of Tarchardina near Sandakan, Sabah, Malaysian Borneo. Yellow crazy ants (Anoplolepis gracilipes) tended Tachardina at this site. Superparasitism (i.e., where more than one progeny is produced per host individual) is frequent. The mean number of Tachardiaephagus somervillei emerging from each parasitized female Tachardina was 2.1 (range 1- 4, N = 30) and up to 5 emergences occurred from each female in Peninsular Malaysia. In contrast, parasitism of Tachardina on Christmas Island was nil; emergence holes were never seen in examination of over 11,000 females over two years (or over the past 15 years of observations of Tachardina on the island).

Page 10 of 44 Environment Protection and Biodiversity Conservation Act 1999

Table 1. Natural enemy assemblages of the yellow lac scale Tachardina aurantiaca on Christmas Island and in Malaysia. + = present, -- = absent. For associates of T. aurantiaca, primary parasitoids oviposit on or in a host and develop within, ultimately killing the host. Hyperparasitoids seek out hosts with primary parasites, oviposit, and develop within the primary parasitoid. Predators feed externally and consume multiple scales. On Christmas Island, search focused on seven areas and examined over 11,000 females and 2000 males of the yellow lac scale. In Malaysia, targeted search for T. aurantiaca on known host plants (e.g., Pongamia pinnata) was frequently used to locate T. aurantiaca. Because the yellow lac scale is so rare across Malaysia, fewer individuals were inspected for parasitization. To assess enemies, host plant twigs with aggregates of T. aurantiaca were first inspected, and then isolated so that emerging parasitoids could be collected for later identification. Individual females of T. aurantiaca were also isolated to collect emergent parasitoids and dissected to determine overall rates of parasitization. On Christmas Island, lac scale predators Eublemma sp. and ?Holcocera sp. were extremely rare.

Association with T. Christmas Species (Family) Malaysia aurantiaca Island

Tachardiaephagus somervillei primary parasitoid -- + Mahdihassan (Encyrtidae)

T. sarawakensis Hayat et al. (Encyrtidae) primary parasitoid -- +

Coccophagus euxanthodes Hayat et al. primary parasitoid -- + (Aphelinidae)

C. tschirchii Mahdihassan (Aphelinidae) primary parasitoid -- +

Coccophagus sp. (Aphelinidae)1 primary parasitoid2 -- +

Promuscidea unfasciativentris Girault hyperparasitoid -- + (Aphelinidae)

Aprostocetus (syn. Tetrastichus) 3 1 hyperparasitoid -- + purpureus Cameron (Eulophidae)

Marietta leopardina Motschulsky 4 primary parasitoid + + (Aphelinidae)

Eublemma sp. (Noctuidae) predator + +

?Holcocera sp. (Blastobasidae) predator + +

1Tentative identification; 2Attack male T. aurantiaca only; 3primary parasitoid of many Coccidae, Diaspididae, Kerriidae, Margarodidae, and Pseudococcidae but known as a hyperparasitoid of C. tschirchii and Tachardiaephagus sp.; 4On Christmas Island and in Malaysia, Marietta leopardina is known only to attack male T. aurantiaca. It has never been observed emerging from female T. aurantiaca. In Southeast Asia, it is also reported as a hyperparasitoid of primary parasitoids of a variety of scale insects.

Supporting Research. Key outcomes from the research conducted by La Trobe University on Christmas Island and in Southeast Asia over the last three years firmly support the concept of indirect biological control (O’Dowd et al. 2012). Briefly,  dietary manipulation experiments in the laboratory, together with studies of stable isotopes in the field, have established a causal link between sugar supply and the dynamics of YCA supercolonies  a large, controlled field experiment has demonstrated an 80% decline in YCA abundance within several weeks of them being denied access to scale insects in the canopies of rainforest trees

Page 11 of 44 Environment Protection and Biodiversity Conservation Act 1999

 while variable, the average contribution made by Tachardina to the honeydew economy of YCA supercolonies is estimated to be 70%  within its native range in Malaysia, Tachardina is rare and patchily distributed, associated with diverse natural enemies including at least five species of primary parasitoids, and suffers high parasitization rates – all attributes consistent with population regulation by natural enemies  host records of Tachardiaephagus somervillei, the most abundant parasitoid of Tachardina across its native range in Malaysia suggest that it is a specialist within the lac scale family, Kerriidae. There are no native lac scale insects on Christmas Island  Tachardiaephagus does not occur on Christmas Island and Tachardina is not currently under effective biological control from the few existing natural enemies.  there is a single ecotype of female Tachardina on Christmas Island, which has been morphologically and genetically matched to populations in Southeast Asia that show evidence of being under control from their primary parasitoids

Approval, Importation & Rearing, and Release & Monitoring of Tachardiaephagus Approval: Currently, there is some uncertainty about the appropriate regulatory framework under which a biological control agent might be considered for introduction to Christmas Island (see Section 2.4, below). However, the framework will almost certainly include host-specificity testing as a necessary precursor to release (Palmer et al. 2010). The purpose of these trials is to test the proposition that Tachardiaephagus is as narrowly specific as the host records suggest, and that the probability of non-target impacts on species of concern is extremely low. Australia has had a long history of conducting host-specificity tests in the area of agent origin, and it has only been relatively recently that these tests have been conducted within Australia by bringing agents into quarantine facilities (Palmer et al. 2010). However, testing within the native range is still desirable because it is safe (no need to import the agent), and it also allows for the possibility of conducting field tests which are seen as more realistic than laboratory trials (Secord and Karieva 1996). In any case, there is no suitable quarantine facility on Christmas Island in which to test Tarchardiaephagus against potential local non-target species, and the cost of constructing one would be prohibitive. Alternatively, Tachardiaephagus could be imported into containment in Australia, but this poses some risk to native kerriid species in Australia, and would present formidable regulatory issues. Instead, host-specificity tests will be conducted within the native distribution of Tachardina aurantiaca at the Forest Research Institute of Malaysia, where La Trobe University has established a research laboratory with the assistance of local co-operators. It is proposed to use “centrifugal host range testing” as the guiding principle to assess the host range of the Tachardiaephagus (Kuhlmann et al 2006, Neumann et al. 2010). In this approach, species other than the known host of the biological control agent are tested with the most closely related species (least phylogenetic distance, and with close similarities in biology and ecology) tested first then less similar species thereafter (“centrifugal principle”). This is outlined in more detail in Attachment 1. Importation & Rearing: Many parasitoids have their own insect natural enemies, known as hyperparasitoids (i.e. parasites of parasites). The importation of a hyperparasite of Tachardieaphagus is inimical to the successful suppression of YCA supercolonies, because it could compromise seriously the capacity of Tachardieaphagus to build up population densities sufficient to control Tachardina. For this reason, great care will be taken to implement standard agent rearing and sanitation techniques to ensure that the founding population of Tachardiaephagus from Malaysia is free of hyperparasitoids and pathogens before importation to Christmas Island.

Page 12 of 44 Environment Protection and Biodiversity Conservation Act 1999

Importation of Tachardiaephagus to the island should conservatively be regarded as the point of release. A production facility will be built on Christmas Island to receive the founder population, and to mass-rear these insects for field-release. This facility will be designed to contain the parasitoids through the rearing phase. The facility will produce Tachardina host plants using existing nursery facilities, and their maintenance in a dedicated glasshouse/greenhouse for the mass rearing of agents. This population will require careful monitoring and maintenance to minimize the selection of laboratory-adapted insects, and to counter the possible susceptibility of Tachardiaephagus to the loss of population heterozygosity. Sex determination in microhymenoptera is usually haplodiploid – males are haploid, females diploid, and heterozygosity at a multi-allelic sex-determining locus is required for femaleness. Inbreeding can lead to a preponderance of homozygous diploids that will either be sterile males, or experience a very high rate of mortality. The infrastructure, personnel and logistic support required for this phase of the biological control program have been outlined in Attachment 2. Release & Monitoring: Protocols will be developed to estimate the impact of Tachardiaephagus on Tachardina and on YCA densities, and criteria will be established to select the number and location of primary release sites and control points, establish the most appropriate timing of the releases according to local site variation in Tachardina phenology, and to determine the rate of spread of Tachardiaephagus from release points. It is premature to give precise details on most aspects of the release and monitoring phase because they will largely depend on the number of insects that can be reared in the production facility, and on the extent and location of YCA supercolonies at the time.

2.2 Alternatives to taking the proposed action This should be a detailed description outlining any feasible alternatives to taking the proposed action (including not taking the action) that were considered but are not proposed (note, this is distinct from any proposed alternatives relating to location, time frames, or activities – see section 2.3).

There are two alternatives to taking the proposed action: 1. Do not take the proposed action. Not taking the proposed action will ensure the management of YCA supercolonies and the amelioration of their impacts will continue to rely on chemical control, mainly the toxicant fipronil. However, the use of fipronil is perpetually reactive, because it relies on continual surveillance, monitoring and control using toxic bait and considerable human resources to keep abreast of supercolony formation. Further, it has not been able to deal with incipient supercolonies where YCA densities are increasing but still not high enough for YCA to monopolise the bait, and thereby minimise non-target impacts. Despite no evidence for the bioaccumulation of fipronil or its metabolites in baited areas (CESAR Consultants 2011), there is still disquiet among all the stakeholders over its continued use but a grudging acceptance of the lack of alternatives (Beeton et al. 2010). For all of these reasons, business-as-usual baiting with fipronil is seen as unsustainable in the long term and is now being considered as a transitional measure until biocontrol can be developed as an alternative (EPBC Referral, Helicopter baiting of the exotic yellow crazy Anoplolepis gracilipes supercolonies on Christmas Island, Indian Ocean, 2012). 2. Develop a biocontrol program for several species of soft scales common in YCA supercolonies, using parasitoids already present on Christmas Island. Coccophagus ceroplastae (Hymenoptera: Aphelinidae) and Encyrtus infelix (Hymenoptera: Encyrtidae), both parasitoids of a variety of coccid scale insects, have been discovered on Christmas Island as part of the supporting research leading up to this referral (O’Dowd et al. 2012). Both were almost certainly introduced inadvertently to the island in the past with importation of plant material with host scale insects. Both species attack a wide variety of soft scales, including four that are important in YCA supercolonies: Coccus hesperidium, C. celatus, Saissetia oleae and S. coffeae.

Page 13 of 44 Environment Protection and Biodiversity Conservation Act 1999

The difficult issues of foreign exploration, host-specificity testing and navigating regulatory frameworks have been obviated by the presence of these parasitoids on the Island. Experience with the efficacy of these agents in dealing with the outbreak of Pulvinaria urbicola on Christmas Island (Neumann et al. 2011) suggests that the current lack of control of soft scales in YCA supercolonies is a result of dispersal limitation of their parasitoids. Dispersal limitation could presumably be overcome by releasing these parasitoids at multiple sites and at multiple times across the island. However, soft scales are estimated to contribute on average just 30% to the honeydew economy of YCA supercolonies, so targeting them alone is less likely to achieve suppression of YCA supercolonies than targeting Tachardina. For this reason, the development of biological control for soft scales using parasitoids already present on the island should be considered as being complementary, rather than as an alternative, to biological control for Tachardina. The goal of the proposed action is to achieve ecologically-effective, cost-effective, and self-sustaining suppression of YCA supercolonies through the importation of a biological control agent specifically for Tachardina. Using these criteria, Alternative 1 is not a viable option because it is not self- sustaining, and Alternative 2 is not a viable option by itself because it is unlikely to achieve supercolony suppression.

2.3 Alternative locations, time frames or activities that form part of the referred action If you have identified that the proposed action includes alternative time frames, locations or activities (in section 1.10) you must complete this section. Describe any alternatives related to the physical location of the action, time frames within which the action is to be taken and alternative methods or activities for undertaking the action. For each alternative location, time frame or activity identified, you must also complete (where relevant) the details in sections 1.2-1.9, 2.4-2.7, 3.3 and 4. Please note, if the action that you propose to take is determined to be a controlled action, any alternative locations, time frames or activities that are identified here may be subject to environmental assessment and a decision on whether to approve the alternative. N/A 2.4 Context, planning framework and state/local government requirements Explain the context in which the action is proposed, including any relevant planning framework at the state and/or local government level (e.g. within scope of a management plan, planning initiative or policy framework). Describe any Commonwealth or state legislation or policies under which approvals are required or will be considered against. Ecological Context. The Expert Working Group’s Final Report (Beeton et al. 2010) provides an ecosystem-wide context for the proposed action. The EWG’s terms of reference included an examination of all threats to Christmas Island’s ecology, biodiversity management and any other issues relating to conservation management of the island and its surrounds. The principal finding of the EWG was that the extremely high biodiversity values of Christmas Island are in a “parlous state” (p. 9), to the extent that Christmas Island and its surrounding seas should be considered for listing as a threatened ecological community under the EPBC Act 1999. Invasion and supercolony formation by YCA was identified as a major contributing factor, and Loss of biodiversity and ecosystem integrity following invasion by the Yellow Crazy Ant (Anoplolepis gracilipes on Christmas Island, Indian Ocean has been listed as a Key Threatening Process under the EPBC Act since 2005. Numerous scientific publications and reports over more than a decade have documented the direct and indirect negative impacts of YCA supercolony formation on various aspects of the island’s ecology (O’Dowd et al. 1999; Green et al. 2001; Marr et al. 2003; O’Dowd et al. 2003; Davis et al. 2008, 2010; Green and O’Dowd 2009; O’Dowd and Green 2010; Green et al. 2011, Smith et al., unpublished results). Planning Framework – The Yellow Crazy Ant Research & Control Program. This 10-year plan was developed by DNP/CINP and CASAP to map out the management of Yellow Crazy Ants from 2007 to 2017. The major objectives of the program include significantly reducing the impact of Yellow Crazy Ants on the biodiversity values of Christmas Island by removing all existing, and halting future, development of supercolonies through:

Page 14 of 44 Environment Protection and Biodiversity Conservation Act 1999

a) development of more target-specific baits which can be used on a more regular and widespread basis. b) moving the baiting control methodology to a predominantly aerial approach. c) research and development of biological control agents to reduce the densities of scale insect mutualists, potentially reducing the main cause of YCA supercolony formation. Part C of this ten-year plan was fully endorsed by two recommendations in the EWG Final Report (Beeton et al. 2010):  Recommendation 9: (High priority) The initial steps taken already to explore biological control of the introduced scale insects be accelerated and biological control trials be started as soon as possible.  Recommendation 29: (High priority) Fundamental investigations continue and be augmented by adaptive management and aspects of Integrated Pest Control experimental work to develop cost-effective methods to break the scale insect - Yellow Crazy Ant mutualistic dependence.

Regulatory Framework – Director of National Parks and the Christmas Island Management Plan. The EPBC Act 1999 outlines the responsibilities of the Director of National Parks (DNP) to protect, conserve and manage biological diversity in Commonwealth reserves. The EPBC Act Regulations provide for the conservation of biological diversity in Commonwealth areas. The EPBC Act requires the Director of National Parks to prepare a management plan for the Park at all times after the first plan for managing the park comes into effect. The third management plan for Christmas Island National Park (2002 - 2009) ceased to be in operation in March 2009 and the fourth management plan is currently in draft (2012) but is likely to be in force in early to mid 2013, before any importation of a biological control agent would occur. Until the fourth Plan comes into operation the Park will be managed in accordance with Section 357 of the EPBC Act. Section 357(1) states: While a management plan is not in operation for a Commonwealth Reserve, the Director must exercise the Director's powers and perform the Director's functions in relation to the reserve or to a zone of the reserve so as to manage the reserve in accordance with: a) the Australian IUCN reserve management principles for the IUCN category to which the reserve or zone has most recently been assigned by: i. a Proclamation made under Subdivision B; or ii. a management plan that was in operation for the reserve (but is no longer); and b) if the Director holds land or seabed included in the reserve under lease - the Directors obligations under the lease. The specific objectives of the draft Christmas Island Management Plan relevant to this referral are: Landscape 4.1.1 Impacts on the values and integrity of the park’s landscape, including visitor experience values will be avoided or minimised. 4.1.7 If the landscape changes in ways that threaten landscape or ecosystem values or otherwise beyond levels of acceptable change, the Director, in consultation with relevant stakeholders, will identify further monitoring requirements and will decide whether protection, rehabilitation or adaptation measures can and will be implemented.

Page 15 of 44 Environment Protection and Biodiversity Conservation Act 1999

Terrestrial vegetation 4.2.3 Subject to risk and required approvals including from the Australian Quarantine and Inspection Service, the Director may introduce, or issue a permit for the introduction of, non- native species for conservation purposes such as the biological control of invasive species 4.2.4 So far as practicable, implement multi-species recovery plans, and relevant threat abatement plans for listed threatened species. Terrestrial animals 4.3.1 Determine priorities for actions to conserve native plant and animal species based on: a) a species’ conservation or biodiversity value, with high priority placed on endemic, keystone and EPBC Act listed and threatened species, and the risks of taking no action b) likelihood that proposed actions will have ecosystem or multiple species benefits c) likelihood that proposed actions will achieve their conservation aims d) cost benefits and effectiveness of implementing proposed actions. 4.3.2 If EPBC Act listed, endemic, keystone or otherwise threatened or significant species are in decline to a level that may threaten their conservation status, the Director will: a) assess the likelihood of mitigating known threats and, if feasible, implement threat mitigation strategies b) if threats are not known, seek to determine (so far as possible) the threats and appropriate mitigation measures c) if threats are not known or not likely to be mitigated for some time, assess the feasibility and effectiveness of implementing interventionist programs, such as captive breeding, that have the long-term aim of conserving the species in their natural environment. 4.3.3 The Director may take actions concerning animal species, including species listed under Part 13 of the EPBC Act, that are otherwise prohibited by the EPBC Act or Regulations where they are necessary to implement this plan, or where they are otherwise necessary for preserving or protecting the park, protecting or conserving biodiversity, or protecting persons or property in the park. 4.3.4 So far as practicable implement, multi-species recovery plans, and relevant threat abatement plans for listed threatened species. 4.3.5 Implement and contribute to on-park and off-park actions for conserving ecosystems and native species, particularly EPBC Act listed and other threatened species including red crabs and other keystone species. 4.3.7 Assess and monitor threats to native species. This will include the risk of invasive species which are currently considered a low threat becoming a greater threat. 4.3.9 Continue to implement the crazy ant control program including monitoring the impact of crazy ants on red crabs and other native species. Ramsar wetlands and other freshwater wetlands 4.4.1 Applicable policies and actions from this plan will be applied to managing The Dales and Hosnies Spring. 4.4.4 Monitor and maintain the ecological character of the Ramsar listed wetlands known as Hosnies Spring and The Dales. This may include mitigating the impacts of potential threats, such as invasive species and unmanaged visitor use.

Page 16 of 44 Environment Protection and Biodiversity Conservation Act 1999

Research and monitoring 8.7.1 Research and monitoring will focus on conserving, managing or restoring the park’s and island’s natural heritage values. 8.7.2 The Director or permitted organisations or individuals may carry out research and monitoring actions concerning native species, including species listed under Part 13 of the EPBC Act. 8.7.3 Permits authorising research and monitoring may be issued for conservation focused research consistent with this plan and with the park’s IUCN category (see Section 3.1). 8.7.4 Persons carrying out research and monitoring in the park will be required to provide the Director with reports on their work in ways requested by the Director. 8.7.6 Where needed, seek independent scientific and other specialist advice, for example through relevant advisory groups, for conserving and managing natural heritage. 8.7.8 Conduct and/or support or encourage conservation management research, monitoring and studies that assist with conserving and managing the park’s and island’s natural heritage, particularly studies on: a) landscape and ecosystem changes and threats including monitoring and management of biodiversity condition b) restoration, recovery and adaptive management of ecosystems and their species including rehabilitated minesites and red crab migration routes/habitat c) populations of threatened and otherwise significant species, including but not limited to seabirds and land crabs (particularly red and robber crabs) and, as needed, invasive species d) identification, quantification and, as needed, management of threatening processes impacting on ecosystems and threatened species. Give priority to invasive species, including the indirect biological control of crazy ants and control of cats and rats 8.7.9 Where possible, provide support for suitably qualified research organisations or individuals conducting natural heritage research consistent with Section 8.7.8.

Regulatory Framework – Commonwealth or state legislation or policies under which approvals are required or will be considered against Any regulatory framework under which the introduction of a biological control agent to Christmas Island would be conducted is currently unclear. However, the Biological Control Act 1984, indicates in Section 4(1) that the responsible Minister, by giving notice in the Gazette, can extend the Act to the external territory of Christmas Island (http://www.comlaw.gov.au/Details/C2008C00315). In the absence of this extension, there is no obvious regulatory framework applicable to these circumstances. For example, the advisory committees (Plant Health Committee [PHC] and the National Biosecurity Committee [NBC]) recommend approvals to the Primary Industry Standing Committee (PISC). A clear approvals process exists for nomination of a biological control target (the yellow lac scale Tachardina aurantiaca [Kerriidae], in this case), but as an external territory Christmas Island appears to be outside of the jurisdiction of both the PHC and NBC. Currently no member on the PHC represents this external island territory; and, even if there were, the island is not within the jurisdiction of the NBC, which approves recommendations made by the PHC. Nevertheless, initial steps have been followed in the Biosecurity Guidelines for the Introduction of Exotic Biological Control Agents for the Control of Weeds and Plant Pests

Page 17 of 44 Environment Protection and Biodiversity Conservation Act 1999

(http://www.daff.gov.au/ba/reviews/biological_control_agents/protocol_for_biological_control_agents) .A nomination of the target species as a candidate for biological control has been produced (see Attachment 4); offshore research in Malaysia on prospective biological control agents is underway; and a rationale for host-specificity testing has been produced (Appendix 1). Approval would be sought to undertake host-specificity testing within the natural range of the biological control agent (Malaysia) rather than in containment on the Australian mainland. It is intended that in the absence of a regulatory framework for introduction of an exotic biological control agent that is applicable to Christmas Island, this referral will act as a catalyst and help provide a basis for the relevant government agencies to identify, in a timely manner, a workable framework under which to consider import approval.

2.5 Environmental impact assessments under Commonwealth, state or territory legislation If you have identified that the proposed action will be or has been subject to a state or territory environmental impact statement (in section 1.11) you must complete this section. Describe any environmental assessment of the relevant impacts of the project that has been, is being, or will be carried out under state or territory legislation. Specify the type and nature of the assessment, the relevant legislation and the current status of any assessments or approvals. Where possible, provide contact details for the state/territory assessment contact officer. Describe or summarise any public consultation undertaken, or to be undertaken, during the assessment. Attach copies of relevant assessment documentation and outcomes of public consultations (if available).

As described above, the regulatory framework under which the introduction of a biological control agent to Christmas Island would be considered cannot be clearly identified. Therefore, the requirement for a formal state or territory environmental impact assessment is also unclear but is considered unlikely, given that the import of exotic species (e.g. plants for propagation) may be permitted (and occurs) under existing biosecurity import arrangements. In lieu of a formal environmental impact assessment, the matters of greatest relevance in the assessment of this Referral are addressed; the host specificity of the proposed biological control agent, and the likelihood of off-target impacts on other scale insects or species of concern. Host Specificity of Tachardiaephagus somervillei. Based on known host records, all Tachardiaephagus species have narrow host ranges and appear to be family specialists, known only to attack scale insects in the Kerriidae, the family to which the yellow lac scale belongs (Table 2; see also O’Dowd et al. 2012).

Table 2. Records of known host families and genera for the primary parasitoid Tachardiaephagus (Encyrtidae). As a genus, Tachardiaephagus has an extremely broad geographic range. With the exception of one probably erroneous host record in Africa, all Tachardiaephagus species appear to be family specialists and restricted to the Kerriidae. For host genera, number of species recorded as hosts is in parentheses. Based on Noyes (2012, Universal Chalcidoidea Database, http://www.nhm.ac.uk/research-curation/research/projects/chalcidoids/database/), except for records for T. somervillei and T. sarawakensis (Hayat et al. 2010. Oriental Insects 44, 23; R.W. Pemberton, pers. comm.; this study)

Species Distribution Recorded hosts

Tachardiaephagus somervillei India, Malaysia, Thailand Kerriidae Kerria spp. (4) Tachardina aurantiaca Tachardina sp. (1)

T. sarawakensis Sarawak Kerriidae Tachardina aurantiaca

T. tachardiae Brunei, China, India, Indonesia, Malaysia, Kerriidae Sri Lanka, Taiwan, Vietnam, Azerbaijan Kerria spp. (6)

Page 18 of 44 Environment Protection and Biodiversity Conservation Act 1999

Laccifer spp. (3) Paratachardina lobata

T. similis Afrotropical, Kerriidae South Africa Tachardina sp. (1) Coccidae ?Ceroplastes eucleae1

T. absonus Afrotropical, Kerriidae South Africa Tachardina spp. (2)

T. communis Afrotropical, Kerriidae South Africa Tachardina spp. (5)

T. gracilis Afrotropical, Kerriidae South Africa Tachardina sp. (1)

1This record is likely to be erroneous (see Prinsloo1983. Entomol. Mem. Dept. Agric. Rep. S. Afr. 60, 26).

Scale insect diversity on Christmas Island. No native or endemic scale insect species have been discovered in 400 hours of search over two years of intensive and extensive searches for scale insect species on Christmas Island (O’Dowd et al. 2012). However, this effort did yield five additional exotic scale insect species previously unknown to the island (Table 3).

Table 3. Scale insects of Christmas Island. It is highly probably that all of these species, with broad host plant ranges and geographic distributions, are exotic to Christmas Island and introduced following human settlement. The yellow lac scale, Tachardina aurantiaca (Kerriidae) and Coccidae (soft scales) are the primary honeydew suppliers to the yellow crazy ant. Scale insects in bold occur commonly in YCA supercolonies. Taxonomy and distribution from Ben-Dov et al. (2012, http://www.sel.barc.usda. gov/scalenet/scalenet.htm).

Honeydew Family and Species Common Name Distribution Producer

Kerriidae (lac scales) Tachardina aurantiaca (Cockerell) Yellow lac scale Oriental Yes *Paratachardina pseudolobata False lobate lac scale Oriental, Nearctic, Neotropical Yes2 (Kondo & Gullan) Coccidae (soft scales) Coccus celatus De Lotto Green coffee scale Afrotropical, Australasia, Oriental Yes *C. hesperidium Linnaeus Brown soft scale Cosmopolitan Yes *Milviscutulus mangiferae (Green) Mango shield scale Cosmopolitan Yes *Ceroplastes ceriferus (Fabricius) Indian wax scale Cosmopolitan Yes *C. destructor Newstead White wax scale Afrotropical, Australasia, Oriental Yes Saissetia oleae (Olivier) Black olive scale Pantropical Yes *S. coffeae (Walker) Hemispherical scale Cosmopolitan Yes **Parasaissetia nigra1 Nigra scale Cosmopolitan Yes **Pulvinaria urbicola Cockerell Urbicola soft scale Pantropical Yes **P. psidii1 Green shield scale Cosmopolitan Yes Monophlebidae (giant scales) Icerya purchasi (Maskell) Cottony cushion scale Cosmopolitan Yes

Cerococcidae (ornate pit scales) **Cerococcus indicus (Maskell) Spiny brown coccid Cosmopolitan Yes

Pseudococcidae (mealybugs) **Dysmicoccus finitimus Williams Asian coconut mealybug Australasia, Oriental Yes **Ferrisia virgata (Cockerell)1 Striped mealybug Cosmopolitan Yes

Diaspididae (armoured scales) Aspidiotus destructor (Signoret) Coconut scale Cosmopolitan No Pseudaulacaspis pentagona White peach scale Cosmopolitan No

Page 19 of 44 Environment Protection and Biodiversity Conservation Act 1999

(Targioni Tozzetti) *Hemiberlesia palmae (Cockerell) Tropical palm scale Cosmopolitan No *Lindingaspis sp. -- -- No **Lepidosaphes sp.1 Oystershell scale -- No

*Record added by Abbott (2004, PhD Thesis, Monash University); **Record added by Neumann et al. (unpublished results); 1Tentative identification; 2P. pseudolobata produces honeydew but ejects it instead of producing droplets that can be collected by ants (Howard 2010, Fl. Entomol. 93, 1).

Likelihood of off-target impacts. All known scale insect species on the island are exotic/invasive, and the proposed biological control agent for Tachardina on Christmas Island appears to be a narrow family specialist (Kerriidae). After intensive survey, no scale insect species of concern have been found on the island, and the probability of any direct non-target effects on any scale insects is negligible. Further, given the co-evolved nature of host-parasitoid interactions, it is extremely unlikely that Tachardiaephagus, as a primary parasitoid wasp, will attack any native Christmas Island insects. Risk of off-target impacts will be addressed by conducting host-specificity tests within the native range of Tachardina in Malaysia where it is safest to conduct these tests.

2.6 Public consultation (including with Indigenous stakeholders) Your referral must include a description of any public consultation that has been, or is being, undertaken. Where Indigenous stakeholders are likely to be affected by your proposed action, your referral should describe any consultations undertaken with Indigenous stakeholders. Identify the relevant stakeholders and the status of consultations at the time of the referral. Where appropriate include copies of documents recording the outcomes of any consultations. Christmas Island has never been settled by Indigenous peoples prior to European settlement. The local community on Christmas Island is very knowledgeable about YCA and its ecological impacts because of public outreach initiatives (community meetings, information brochures and stakeholder meetings) carried out by the DNP and La Trobe University over a number of years The local community is also very aware of the need to suppress supercolonies using toxic bait, especially because of the high community profile of the issue generated by three helicopter baiting campaigns, and the recent (September 2012) ground baiting of YCA supercolonies behind the Kampong, a major residential area. The concept of a biological control program for the suppression of YCA supercolonies was first introduced to the community through public presentations and meetings in 2009. University researchers Green and O’Dowd gave repeat presentations at the Malay Club, Sports and Recreation Centre, the Chinese Literary Association, the Christmas Island District High School and to Phosphates Resources Limited, in an effort to inform a wide cross section of the community of the proposal. The proposal was welcomed insofar as it presents an alternative means of suppressing supercolonies. All community groups appreciated the potential risks of introducing another non-native species to Christmas Island, but were reassured that risk assessment and strict regulatory oversight would occur prior to any importation and release of any biological control agent. Further public meetings will be held closer to the time of importation and release. In addition, several key island based stakeholders, especially the Department of Regional Australia, Local Government, Arts and Sport (DRALGAS) are aware of and support (in-principle) the biological control research. This is reflected in the Australian Government’s response (http://www.environment.gov.au/parks/publications/christmas/ewg-response.html) to the Expert Working Groups final report (Beeton et al. 2010), particularly recommendation nine that: “the initial steps taken already to explore biological control of the introduced scale insects be accelerated and biological control trials be started as soon as possible”.

2.7 A staged development or component of a larger project If you have identified that the proposed action is a component of a larger action (in section 1.12) you must complete this section. Provide information about the larger action and details of any interdependency between the stages/components and the larger action. You may also provide justification as to why you believe it is reasonable for the referred action to be

Page 20 of 44 Environment Protection and Biodiversity Conservation Act 1999

considered separately from the larger proposal (eg. the referred action is ‘stand-alone’ and viable in its own right, there are separate responsibilities for component actions or approvals have been split in a similar way at the state or local government levels).

The proposed action is part of a larger program of integrated pest management designed to ameliorate the impacts of YCA invasion by suppressing supercolony formation (see section 2.4 above, Planning Framework). The current backbone of the program is surveillance and monitoring to detect and map supercolonies, followed by the use of toxic bait, distributed by hand on the ground or by helicopter, to suppress them. The bait is AntOff®, with the active ingredient Fipronil at 0.01 g kg-1 in a fishmeal protein matrix, broadcast at 4 kg ha-1. Several alternative chemical approaches, including different neurotoxins and insect growth regulators (IGRs), have also been trialled but so far without success. New trials using IGRs were commenced during the aerial baiting campaign of September 2012 but the outcome of those trials will not be known until early to mid-2013.

3 Description of environment & likely impacts

3.1 Matters of national environmental significance Describe the affected area and the likely impacts of the proposal, emphasising the relevant matters protected by the EPBC Act. Refer to relevant maps as appropriate. The interactive map tool can help determine whether matters of national environmental significance or other matters protected by the EPBC Act are likely to occur in your area of interest.

Your assessment of likely impacts should refer to the following resources (available from the Department’s web site):  specific values of individual World Heritage properties and National Heritage places and the ecological character of Ramsar wetlands;  profiles of relevant species/communities (where available), that will assist in the identification of whether there is likely to be a significant impact on them if the proposal proceeds;  Significant Impact Guidelines 1.1 – Matters of National Environmental Significance; and  associated sectoral and species policy statements available on the web site, as relevant.

Your assessment of likely impacts should consider whether a bioregional plan is relevant to your proposal. The Minister has prepared four marine bioregional plans (MBP) in accordance with section 176. It is likely that the MBP’s will be more commonly relevant where listed threatened species, listed migratory species or a Commonwealth marine area is considered.

Note that even if your proposal will not be taken in a World Heritage area, Ramsar wetland, Commonwealth marine area, the Great Barrier Reef Marine Park or on Commonwealth land, it could still impact upon these areas (for example, through downstream impacts). Consideration of likely impacts should include both direct and indirect impacts.

3.1 (a) World Heritage Properties

Description N/A. There are no World Heritage Properties on Christmas Island

Nature and extent of likely impact Address any impacts on the World Heritage values of any World Heritage property. N/A. There are no World Heritage Properties on Christmas Island

Page 21 of 44 Environment Protection and Biodiversity Conservation Act 1999

3.1 (b) National Heritage Places

Description (from Draft Christmas Island National Park Management Plan, Commonwealth of Australia 2012). Christmas Island is not listed as a National Heritage Place but is listed on the Commonwealth Heritage register for both natural and cultural reasons (Commonwealth of Australia 2012). Christmas Island has a number of unique ecological characteristics, as well as other features common to other oceanic islands. Characteristics of Christmas Island’s ecosystems include: fewer species than comparable areas of land on continents or islands near continents but with a relatively higher proportion of endemic terrestrial species; evolutionary isolation for thousands or millions of years until recent human arrival; native species that have evolved with few competitors; and many species with small population sizes. These characteristics make the island’s ecosystems and species of high conservation value but also highly vulnerable to environmental change, particularly from invasive species, climate change, and habitat clearing, fragmentation and degradation. Christmas Island also has nine cultural sites on the register. Eight of the nine sites are located within settled areas of the island, and the other site (South Point Settlement remains) is well away from settled areas. Nature and extent of likely impact Address any impacts on the National Heritage values of any National Heritage place. The likelihood of negative impacts on species of concern from the introduction of a biological control agent is rated as negligible (see Section 2.5); therefore, the likelihood of negative impacts on Commonwealth Heritage Places is also rated as negligible. The likelihood of negative impacts on cultural sites is nil.

3.1 (c) Wetlands of International Importance (declared Ramsar wetlands)

Description (extracted from the Draft Christmas Island National Park Management Plan, Commonwealth of Australia 2012)

The Dales (RS1225). The Dales Ramsar site includes a series of seven dales, three of which support permanent springs and four support intermittent streams. The Dales are surrounded predominantly by semi-deciduous forest. On the seaward side at the edge of the shore terrace there is a line of coastal shrubland which merges with sea cliffs and rocky marine shores. The site extends seaward 50 metres and includes part of a narrow, shallow, sloping reef. Mixed amongst the terrestrial and marine environments is a range of karst features, highly representative of the environment of Christmas Island. The combination of this variety of habitats and the presence of permanent surface water supports a wide diversity of endemic and threatened species. The site hosts part of the annual red crab migration and provides critical habitat for blue crabs as well as other land crabs. The Dales supports a diverse community of tree species and epiphytes. At Hugh’s Dale, and in parts of Anderson Dale and Sydney’s Dale, there are mono-specific stands of Tahitian Chestnut, Inocarpus fagifer, and the rare epiphytic Ribbon Fern, Ophioglossum pendulum. The endemic Arenga Palm, Arenga listeri, and endemic Ridley’s Orchid, Brachypeza archytas, are common in The Dales. Terminalia catappa grows to an unusual size on Christmas Island and several large specimens occur in The Dales. A number of endemic fauna occur within The Dales including the Abbott’s Booby and several land bird species. The endemic freshwater fish, the Brown Gudgeon, Eleotris fusca, has also been recently sighted within The Dales. Hosnies Spring (RS512). The Hosnies Spring Ramsar site is an area of permanent, shallow freshwater wetland, fed by a natural spring system located approximately 30 metres above sea level

Page 22 of 44 Environment Protection and Biodiversity Conservation Act 1999

and 120 metres inland of the seaward cliff. The wetland is covered by a stand of mangroves including Bruguiera gymnorhiza and B. sexangula estimated to be 120,000 years old. The margins of the wetland are well defined, with limestone cliffs to the north and west and a sharp transition to hibiscus and pandanus to the south. The area that surrounds the wetland site is predominantly rainforest, characterised by a 20 to 30 metre tall canopy of evergreen and deciduous tree species such as Pisonia grandis and Barringtonia racemosa and a conspicuous lack of herb and shrub layers. There is a narrow band of coastal scrub with hardy species such as Scaevola taccada at the seaward margin of the shore terrace, with an unvegetated area of limestone pinnacles at the top of the sea cliff. The cliff descends some 17 metres almost vertically to the rocky marine shore below. The site extends 50 metres seaward of the low water mark and includes areas of shallow coral reef. Hosnies Spring is remarkable for a number of reasons. First, it is one the few permanent freshwater areas on Christmas Island. Second, the mangroves occur at an elevation not recorded anywhere else in the world. Third, the age of the mangrove stand is extraordinary and finally, the individual trees are very large. The site also supports endemic and other significant fauna species including land crabs (in particular red, robber and blue crabs), sea and land birds and the Christmas Island flying fox. In 2010, Ecological Character Descriptions for The Dales and Hosnies Spring were prepared under the National Framework and Guidance for Describing the Ecological Character of Australia’s Ramsar Wetlands (Hale and Butcher 2010). These descriptions form a baseline reference to help monitor and maintain the ecological character of The Dales and Hosnies Spring.

Nature and extent of likely impact Address any impacts on the ecological character of any Ramsar wetlands. Because the biological control agent is a host-specific, parasitoid insect, there are no foreseeable negative impacts associated with its introduction on the conservation values of either Ramsar Wetland. It is anticipated that there will be positive impacts on Ramsar sites, because of the likely reduction of YCA densities at these sites.

Page 23 of 44 Environment Protection and Biodiversity Conservation Act 1999

3.1 (d) Listed threatened species and ecological communities

Description

EPBC Act - Listed threatened species occurring in Christmas Island National Park

Common name Scientific name Status MAMMALS

Christmas Island shrew Crocidura attenuata trichura EN Humpback whale Megaptera novaeangliae VU Christmas Island pipistrelle Pipistrellus murrayi CR Maclear's rat Rattus macleari EX Bulldog rat Rattus nativitatis EX BIRDS Abbott's booby Papasula abbotti EN Christmas Island emerald dove Chalcophaps indica natalis EN Christmas Island frigatebird Fregata andrewsi VU Christmas Island goshawk Accipiter fasciatus natalis EN Christmas Island hawk-owl Ninox natalis VU Christmas Island thrush Turdus poliocephalus erythropleurus EN REPTILES Green turtle Chelonia mydas VU Hawksbill turtle Eretmochelys imbricata VU Tree gecko Lepidodactylus listeri VU Pink blind snake Typhlops exocoeti VU FISH Whale shark Rhincodon typus VU VASCULAR PLANTS Christmas Island spleenwort Asplenium listeri CR A fern Pneumatopteris truncata CR A fern Tectaria devexa var. minor EN

EN: endangered; EX: extinct; CR: critically endangered; VU: vulnerable

Nature and extent of likely impact Address any impacts on the members of any listened threatened species (except a conservation dependent species) or any threatened ecological community, or their habitat. Because the proposed biological agent is a primary parasitoid of one particular family of introduced scale insects, there are no foreseeable negative impacts of its introduction on any EPBC-Listed species. It is anticipated that there will be positive impacts on any EPBC-Listed (terrestrial) species, because of the likely reduction of YCA densities in their habitats.

Page 24 of 44 Environment Protection and Biodiversity Conservation Act 1999

3.1 (e) Listed migratory species

Description There are 68 bird species classed as migratory (or vagrant) detected on Christmas Island and five marine species including reptiles, mammals and one fish (see table in section 3.1d) that are recognised by EPBC, JAMBA and/or CAMBA and the Bonn Convention.

EPBC Act EPBC Act Common name Scientific name marine migratory JAMBA CAMBA ROKAMBA BONN MAMMALS Megaptera   Humpback whale novaeangliae Stenella longirostris (E. Long-snouted spinner tropical Pacific/SE   dolphin Asian pops) BIRDS Papasula abbotti =    Abbott’s booby Sula abbotti Antarctic prion Pachyptila desolata  Pelecanus  Australian pelican conspicillatus Australian pratincole Stiltia isabella  Stercorarius    Arctic jaeger parasiticus Baillon’s crake Porzana pusilla  Barn swallow Hirundo rustica      Bar-tailed godwit Limosa lapponica       Himanotopus  Black-winged stilt himanotopus Brown booby Sula leucogaster      Brown goshawk Accipiter fasciatus  Bulwer’s petrel Bulweria bulwerii  Ardea ibis = Bubulcus     Cattle egret ibis Christmas Island    frigatebird Fregata andrewsi Common greenshank,       greenshank Tringa nebularia Eudynamys  Common koel scolopacea Common noddy Anous stolidus     Actitis hypoleucos       Common sandpiper =Tringa hypoleucos Common tern Sterna hirundo      Crested tern Sterna bergii  Curlew sandpiper Calidris ferruginea      Dollarbird Eurystomus orientalis  Eastern reef egret Egretta sacra   Fork-tailed swift Apus pacificus      Garganey Anas querquedula       Glossy ibis Plegadis falcinellus     Ardea alba = Egretta     Great egret alba Great frigatebird Fregata minor     Great knot Calidris tenuirostris      Great skua Catharacta skua 

Page 25 of 44 Environment Protection and Biodiversity Conservation Act 1999

Charadrius      Greater sand plover leschenaultii Grey phalarope Phalaropus fulicarius    Grey plover Pluvialis squaterola      Grey wagtail Motacilla cinerea     Grey-tailed tattler Heteroscelus brevipes       Least frigatebird Fregata ariel      Little curlew Numenius minutus       Little egret Egretta garzetta  Little ringed plover Charadrius dubius   Little tern Sterna albifrons      Gorsachius  Malay night heron melanolophus Marsh sandpiper, little       greenshank Tringa stagnatilis Mongolian plover = lesser       sand plover Charadrius mongolus Nankeen kestrel Falco cenchroides  Nankeen night heron Nycticorax caledonicus  Oriental cuckoo Cuculus saturatus     

Charadrius veredus =      Oriental plover C. asiaticus veredus Oriental pratincole Glareola maldivarum      Oriental reed-warbler Acrocephalus orientalis     Pacific (= lesser) golden       plover Pluvialis fulva Pallid cuckoo Cuculus pallidus  Pied imperial-pigeon Ducula bicolor  Pin-tailed snipe Gallinago stenura       Red-footed booby Sula sula     Red-necked phalarope Phalaropus lobatus       Red-necked stint Calidris ruficollis       Red-rumped swallow Hirundo daurica    Red-tailed tropicbird Phaethon rubricauda  Anthus  Richard’s pipit novaeseelandiae Ruddy turnstone Arenaria interpres       Sacred kingfisher Todiramphus sanctus  Calidris alba =       Sanderling Crocethia alba Sharp-tailed sandpiper Calidris acuminata       Sooty tern Sterna fuscatai  Swinhoe’s snipe Gallinago megala       Terek sandpiper Xenus cinereus       Tree martin Hirundo nigricans  Wedge-tailed shearwater Puffinus pacificus    Whimbrel Numenius phaeopus       Whiskered tern Chlidonias hybridus  White tern Gygis alba  White-tailed tropicbird Phaethon lepturus     White wagtail Motacilla alba    White-bellied sea-eagle Haliaeetus leucogaster   

White-throated needletail Hirundapus      = Spine-tailed swift caudacutus

Page 26 of 44 Environment Protection and Biodiversity Conservation Act 1999

White-winged black tern Chlidonias leucopterus      Wood sandpiper Tringa glareola       Yellow wagtail Motacilla flava      REPTILES Green turtle Chelonia mydas    Hawksbill turtle Eretmochelys imbricata    Yellow-bellied sea snake Pelamis platurus  FISH Bluestripe pipefish Doryramphus excisus  Corythoichthys  Fijian pipefish amplexus Solenostomus  Ornate ghost pipefish paradoxus Redstripe pipefish Doryramphus baldwini  Corythoichthys  Reef-top pipefish haematopterus Solenostomus  Robust ghost pipefish cyanopterus Roughridge pipefish Cosmocampus banneri  Schultz’s pipefish Corythoichthys schultzi  Sculptured pipefish Choeroichthys sculptus  Choeroichthys  Short-bodied pipefish brachysoma Thorn-tailed pipefish = Micrognathus  pygmy pipefish brevirostris pygmaeus Whale shark Rhincodon typus   Corythoichthys  Yellow-banded pipefish flavofasciatus JAMBA – Japan-Australia Migratory Bird Agreement; CAMBA – China-Australia Migratory Bird Agreement; ROKAMBA – Korea- Australia Migratory Bird Agreement; Bonn – Bonn Convention

Nature and extent of likely impact Address any impacts on the members of any listed migratory species, or their habitat. Because the proposed biological agent is a primary parasitoid with hosts within one particular family of introduced scale insects, there are no foreseeable negative impacts of its introduction on any listed migratory species. It is anticipated that there will be positive impacts on EPBC-Listed migratory (terrestrial) species, because of the likely reduction of YCA densities in their habitats.

3.1 (f) Commonwealth marine area (If the action is in the Commonwealth marine area, complete 3.2(c) instead. This section is for actions taken outside the Commonwealth marine area that may have impacts on that area.) Description N/A. The proposed action will not be taken in a marine area that lies outside the Commonwealth marine area

Nature and extent of likely impact Address any impacts on any part of the environment in the Commonwealth marine area. N/A. The proposed action will not be taken in a marine area that lies outside the Commonwealth marine area

Page 27 of 44 Environment Protection and Biodiversity Conservation Act 1999

3.1 (g) Commonwealth land (If the action is on Commonwealth land, complete 3.2(d) instead. This section is for actions taken outside Commonwealth land that may have impacts on that land.) Description If the action will affect Commonwealth land also describe the more general environment. The Policy Statement titled Significant Impact Guidelines 1.2 - Actions on, or impacting upon, Commonwealth land, and actions by Commonwealth agencies provides further details on the type of information needed. If applicable, identify any potential impacts from actions taken outside the Australian jurisdiction on the environment in a Commonwealth Heritage Place overseas. N/A. The proposed action will not be taken in a terrestrial area that lies outside Commonwealth land

Nature and extent of likely impact Address any impacts on any part of the environment in the Commonwealth land. Your assessment of impacts should refer to the Significant Impact Guidelines 1.2 - Actions on, or impacting upon, Commonwealth land, and actions by Commonwealth agencies and specifically address impacts on:  ecosystems and their constituent parts, including people and communities;  natural and physical resources;  the qualities and characteristics of locations, places and areas;  the heritage values of places; and  the social, economic and cultural aspects of the above things. N/A. The proposed action will not be taken in a terrestrial area that lies outside Commonwealth land

3.1 (h) The Great Barrier Reef Marine Park

Description

N/A. The proposed action will take place on Christmas Island, several thousand kilometres away from the GBRMP

Nature and extent of likely impact Address any impacts on any part of the environment of the Great Barrier Reef Marine Park. Note: If your action occurs in the Great Barrier Reef Marine Park you may also require permission under the Great Barrier Reef Marine Park Act 1975 (GBRMP Act). If so, section 37AB of the GBRMP Act provides that your referral under the EPBC Act is deemed to be an application under the GBRMP Act and Regulations for necessary permissions and a single integrated process will generally apply. Further information is available at www.gbrmpa.gov.au N/A. The proposed action will take place on Christmas Island, several thousand kilometres away from the GBRMP

3.2 Nuclear actions, actions taken by the Commonwealth (or Commonwealth agency), actions taken in a Commonwealth marine area, actions taken on Commonwealth land, or actions taken in the Great Barrier Reef Marine Park You must describe the nature and extent of likely impacts (both direct & indirect) on the whole environment if your project:  is a nuclear action;  will be taken by the Commonwealth or a Commonwealth agency;  will be taken in a Commonwealth marine area;  will be taken on Commonwealth land; or  will be taken in the Great Barrier Reef marine Park.

Your assessment of impacts should refer to the Significant Impact Guidelines 1.2 - Actions on, or impacting upon, Commonwealth land, and actions by Commonwealth agencies and specifically address impacts on:  ecosystems and their constituent parts, including people and communities;  natural and physical resources;  the qualities and characteristics of locations, places and areas;  the heritage values of places; and  the social, economic and cultural aspects of the above things.

Page 28 of 44 Environment Protection and Biodiversity Conservation Act 1999

3.2 (a) Is the proposed action a nuclear action? No

Yes (provide details below) If yes, nature & extent of likely impact on the whole environment

3.2 (b) Is the proposed action to be taken by the No Commonwealth or a Commonwealth agency? Yes (provide details below)

If yes, nature & extent of likely impact on the whole environment Researchers from La Trobe University will undertake the proposed action under contract to the Director of National Parks. As outlined in detail in Sections 2.5, 3.1(b-e) and Section 5, there are no foreseeable negative impacts on the environment.

3.2 (c) Is the proposed action to be taken in a Commonwealth marine area? No

Yes (provide details below) If yes, nature & extent of likely impact on the whole environment (in addition to 3.1(f))

3.2 (d) Is the proposed action to be taken on No Commonwealth land? Yes (provide details below)

If yes, nature & extent of likely impact on the whole environment (in addition to 3.1(g)) As outlined in detail in Sections 2.5, 3.1(b-e) and Section 5, there are no foreseeable negative impacts on the environment.

3.2 (e) Is the proposed action to be taken in the Great Barrier Reef Marine Park? No

Yes (provide details below) If yes, nature & extent of likely impact on the whole environment (in addition to 3.1(h))

Page 29 of 44 Environment Protection and Biodiversity Conservation Act 1999

3.3 Other important features of the environment Provide a description of the project area and the affected area, including information about the following features (where relevant to the project area and/or affected area, and to the extent not otherwise addressed above). If at Section 2.3 you identified any alternative locations, time frames or activities for your proposed action, you must complete each of the details below (where relevant) for each alternative identified.

3.3 (a) Flora and fauna Christmas Island’s isolation, climate and the influence of land crabs have resulted in the development of distinct tropical rainforest vegetation, including some species that have evolved to be taller and larger than examples of the same species found elsewhere. There are approximately 213 native plant species on Christmas Island, of which 17 are endemic (Claussen 2005). Holmes and Holmes (2002) list six endangered species, three vulnerable species, thirty three rare species and eleven as poorly known. A number of these species are listed as threatened under the EPBC Act. The island’s terrestrial fauna is dominated by ecologically important and diverse land crabs. Red crabs (Gecarcoidea natalis) are the most numerous, and they act as keystone species because by regulating seedling recruitment, they influence structure and species composition of the island’s rainforest vegetation (Green et al. 1997, 2008). Red crabs also regulate the entry and spread of some invasive species in rainforest on the island (Lake and O’Dowd 1991, Green et al. 2011). The island also supports the world’s largest remaining population of the Robber Crab, Birgus latro (Drew et al. 2010). Christmas Island is classified an Endemic Bird Area by BirdLife International, with nine nesting varieties /species of seabirds and around 80,000 resident birds. More than 100 seabird species have been recorded, including two endemic seabird species and one endemic subspecies that breed on the island. The Abbott's Booby Papasula abbotti (Listed as Endangered), Christmas Island Frigate Bird Fregata andrewsi, and the Golden Bosun Phaethon lepturus fulvus only nest on Christmas Island. There are seven endemic land birds, including the threatened Christmas Island Hawk-owl Ninox natalis, Christmas Island Thrush Turdus poliocaphalus erythropleurus, Christmas Island Goshawk Accipiter fasciatus natalis, and Christmas Island Emerald Dove Chalcophaps indica natalis. The island is also important for other terrestrial fauna species. Five of the six recorded native reptile species are endemic; the Blue-tailed Skink Cryptoblepharus egeriae, Lister’s Gecko Lepidodactylus listeri, the Forest Skink Emoia nativitatis, the Giant Gecko Cyrtodactylus sadlieri, the Coastal Skink Emoia atrocostata, and the Christmas Island Blind Snake Typhlops exocoeti. Lister’s gecko and the Christmas Island Blind Snake are listed under the EPBC Act as vulnerable. The range and abundance of Blue-tailed Skinks, Forest Skinks, Costal Skinks and Lister’s Geckos have dramatically contracted over the past decade and all are currently on the brink of extinction in the wild (Smith et al. unpublished results). Five native endemic mammals have been recorded. The Bulldog Rat Rattus nativitatus, and Maclear’s Rat Rattus macleari, are extinct and the Christmas Island Shrew, Crocidura attenuata trichura, is likely to be extinct. Murray’s Pipistrelle Bat Pipistrellus murrayi was once widespread but is now presumed extinct (Beeton et al. 2010). Yellow Crazy Ants are considered a threat to both the Pipistrelle Bat (Schultz and Lumsden 2004) and the Christmas Island Shrew (Schultz 2004).

3.3 (b) Hydrology, including water flows Most rainforest on Christmas Island occurs over ancient limestone reefs of marine origin, which sit perched on basaltic bedrock. The soils are extremely porous, and Christmas Island has little flowing water. The few streams and soaks occur on the eastern and western coasts of the island, where the interface of the limestone and basalt occurs above sea level.

Page 30 of 44 Environment Protection and Biodiversity Conservation Act 1999

3.3 (c) Soil and Vegetation characteristics The Christmas Island environment can be classified into 12 broad environments: 1) marine – ocean water, sand flats, caves, coral reefs and walls, 2) shoreline rock platforms, 3) beaches, 4) sea cliffs, 5) terrace forest, 6) shallow soil rainforest on the higher terraces, 7) limestone scree slopes and pinnacles, cliffs 8) deeper plateau and terrace soils rainforest, 9) mangrove forest, 10) perennially wet areas, 11) karst, comprising caves, overhangs, rock crevices, and sinkholes, and 12) mining field. The biological control agent will be released at multiple locations in environments 5, 6 and 8, but will hopefully spread to 7, 9, 10 and 12 as well.

3.3 (d) Outstanding natural features There are several mapped caves in karst environments on the island. These will not be affected by the introduction of a biological agent.

3.3 (e) Remnant native vegetation (from Draft Christmas Island Management Plan, Commonwealth of Australia 2012) The island’s remoteness, climate and the influence of land crabs have resulted in the development of distinct tropical rainforest vegetation, including some species that have evolved to be taller and larger than examples of the same species found elsewhere. There are at least 17 endemic plant species, including a rare fern Asplenium listeri, a tall treelike pandanus Pandanus elatus, and a palm Arenga listeri. There are also relict populations of mangrove species and cycads which have been left isolated by the island’s tectonic uplift. A number of these species are listed as threatened under the EPBC Act. Many of the world’s tropical rainforests, particularly on islands in the Indian Ocean, the Pacific Ocean and South-East Asian regions, are under threat from human activities. This makes the park’s and the island’s rainforest ecosystems of increasingly significant conservation value. Vegetation and species of significance will not be adversely affected by the introduction of the biological control agent.

3.3 (f) Gradient (or depth range if action is to be taken in a marine area) N/A

3.3 (g) Current state of the environment Include information about the extent of erosion, whether the area is infested with weeds or feral animals and whether the area is covered by native vegetation or crops.

Page 31 of 44 Environment Protection and Biodiversity Conservation Act 1999

N/A

3.3 (h) Commonwealth Heritage Places or other places recognised as having heritage values As for Section 3.1b

3.3 (i) Indigenous heritage values Christmas Island was never been settled by Indigenous peoples prior to European settlement.

3.3 (j) Other important or unique values of the environment Describe any other key features of the environment affected by, or in proximity to the proposed action (for example, any national parks, conservation reserves, wetlands of national significance etc). As for Section 1.4

3.3 (k) Tenure of the action area (eg freehold, leasehold) Commonwealth Land

3.3 (l) Existing land/marine uses of area National Park and Crown Land

3.3 (m) Any proposed land/marine uses of area There are no additional proposed land uses for the area

Page 32 of 44 Environment Protection and Biodiversity Conservation Act 1999

4 Measures to avoid or reduce impacts

Note: If you have identified alternatives in relation to location, time frames or activities for the proposed action at Section 2.3 you will need to complete this section in relation to each of the alternatives identified.

Provide a description of measures that will be implemented to avoid, reduce, manage or offset any relevant impacts of the action. Include, if appropriate, any relevant reports or technical advice relating to the feasibility and effectiveness of the proposed measures.

For any measures intended to avoid or mitigate significant impacts on matters protected under the EPBC Act, specify:  what the measure is,  how the measure is expected to be effective, and  the time frame or workplan for the measure.

Examples of relevant measures to avoid or reduce impacts may include the timing of works, avoidance of important habitat, specific design measures, or adoption of specific work practices.

Provide information about the level of commitment by the person proposing to take the action to implement the proposed mitigation measures. For example, if the measures are preliminary suggestions only that have not been fully researched, or are dependent on a third party’s agreement (e.g. council or landowner), you should state that, that is the case.

Note, the Australian Government Environment Minister may decide that a proposed action is not likely to have significant impacts on a protected matter, as long as the action is taken in a particular manner (section 77A of the EPBC Act). The particular manner of taking the action may avoid or reduce certain impacts, in such a way that those impacts will not be ‘significant’. More detail is provided on the Department’s web site.

For the Minister to make such a decision (under section 77A), the proposed measures to avoid or reduce impacts must:  clearly form part of the referred action (eg be identified in the referral and fall within the responsibility of the person proposing to take the action),  be must be clear, unambiguous, and provide certainty in relation to reducing or avoiding impacts on the matters protected, and  must be realistic and practical in terms of reporting, auditing and enforcement.

More general commitments (eg preparation of management plans or monitoring) and measures aimed at providing environmental offsets, compensation or off-site benefits CANNOT be taken into account in making the initial decision about whether the proposal is likely to have a significant impact on a matter protected under the EPBC Act. (But those commitments may be relevant at the later assessment and approval stages, including the appropriate level of assessment, if your proposal proceeds to these stages).

Many parasitoid wasps are constrained by their evolution to be the enemies of a narrow range of hosts. In Section 2.5, it was demonstrated that the proposed biological control agent Tachardiaephagus somervillei is almost certainly a highly specialised enemy of scale insects in the family Kerriidae; all Tachardiaephagus species have a narrow host range and appear to be family specialists, known only to attack lac scale insects in the Kerriidae, the family to which the yellow lac scale Tachardina belongs (Table 2). Just two kerriid species occur on Christmas Island, the target Tachardina aurantiaca, and the false lobate lac scale Paratachardina pseudolobata. Both are introduced, exotic and invasive on Christmas Island. There is one host record of a closely related parasitoid Tachardiaephagus tachardiae from another Paratachardina species, P. lobata in India (Table 2). Even if the host range of Tachardiaephagus somervillei did extend to invasive Paratachardina on the island, this would have no negative consequences. Based on the research conducted by La Trobe University the likelihood that Tachardiaephagus will have non-target impacts on any other species, including other scale insects, is considered remote. This view is supported by the Crazy Ant Scientific Advisory Panel (CASAP) (see Attachment 3). However, despite some spectacular successes elsewhere, the history of biocontrol has also been marred by a few failures when the biological control agent itself proved to be problematic. Most of these instances were not (unlike this proposal) based on pre-release research, involved the introduction of generalist or oligophagous predators prior to the establishment of regulatory frameworks for biological control introductions, or failed to consider the spread of the biological

Page 33 of 44 Environment Protection and Biodiversity Conservation Act 1999

control agent beyond the target area. In most cases, once a biological control agent has been released there is usually little that can be done to avoid or reduce unintended impacts. Clearly, such situations are best avoided entirely by making as sure as possible that Tachardiaephagus is as narrowly specific as the host records suggest (Table 2). Host specificity testing is required by Australia under the Biological Control Act 1984. However, no such regulatory framework has been identified that is currently applicable to the external territory of Christmas Island (end of Section 2.4). In the interim, La Trobe University will continue with plans to conduct pre-importation host-specificity trials with the assistance of cooperators at the Forest Research Institute of Malaysia at Kuala Lumpur and at Sarawak Forestry in Semenggoh.

Page 34 of 44 Environment Protection and Biodiversity Conservation Act 1999

5 Conclusion on the likelihood of significant impacts Identify whether or not you believe the action is a controlled action (ie. whether you think that significant impacts on the matters protected under Part 3 of the EPBC Act are likely) and the reasons why.

5.1 Do you THINK your proposed action is a controlled action?

No, complete section 5.2

Yes, complete section 5.3

5.2 Proposed action IS NOT a controlled action. Specify the key reasons why you think the proposed action is NOT LIKELY to have significant impacts on a matter protected under the EPBC Act.

 The proposed biological control agent is almost certainly host-specific on one target invasive species  Non-target impacts, although highly unlikely, will almost certainly be restricted to one other exotic and invasive species, in the same subfamily as the target.

5.3 Proposed action IS a controlled action Type ‘x’ in the box for the matter(s) protected under the EPBC Act that you think are likely to be significantly impacted. (The ‘sections’ identified below are the relevant sections of the EPBC Act.)

Matters likely to be impacted

World Heritage values (sections 12 and 15A)

National Heritage places (sections 15B and 15C)

Wetlands of international importance (sections 16 and 17B)

Listed threatened species and communities (sections 18 and 18A)

Listed migratory species (sections 20 and 20A)

Protection of the environment from nuclear actions (sections 21 and 22A)

Commonwealth marine environment (sections 23 and 24A)

Great Barrier Reef Marine Park (sections 24B and 24C)

Protection of the environment from actions involving Commonwealth land (sections 26 and 27A)

Protection of the environment from Commonwealth actions (section 28)

Commonwealth Heritage places overseas (sections 27B and 27C)

Specify the key reasons why you think the proposed action is likely to have a significant adverse impact on the matters identified above.

Page 35 of 44 Environment Protection and Biodiversity Conservation Act 1999

6 Environmental record of the responsible party NOTE: If a decision is made that a proposal needs approval under the EPBC Act, the Environment Minister will also decide the assessment approach. The EPBC Regulations provide for the environmental history of the party proposing to take the action to be taken into account when deciding the assessment approach.

Yes No 6.1 Does the party taking the action have a satisfactory record of responsible environmental management?

Provide details The Commonwealth Director of National Parks is the responsible party for this action. The DNP manages several Commonwealth reserves on behalf of the Australian Government, in accordance with the EPBC Act and relevant park Management Plans. These reserves include the Christmas Island National Park, and the World Heritage listed Kakadu and Uluru Kata- Tjuta National Parks. On matters pertaining to the management of the YCA invasion on Christmas Island, the DNP considers advice from the Crazy Ant Scientific Advisory Panel, comprised of researchers with expertise in invasive species biology including invasive ants, expertise with the local Island environment, and expertise in policy development, implementation and ecosystem management. The on-ground research and implementation of the biocontrol program will continue to be the responsibility of La Trobe University, funded by the DNP. The La Trobe team has a very long history of research on the island, and two members were part of the team with DNP/CINP that won the 2003 Banksia Award for environmental management.

6.2 Has either (a) the party proposing to take the action, or (b) if a permit has been applied for in relation to the action, the person making the application - ever been subject to any proceedings under a Commonwealth, State or Territory law for the protection of the environment or the conservation and sustainable use of natural resources?

If yes, provide details

6.3 If the party taking the action is a corporation, will the action be taken in N/A accordance with the corporation’s environmental policy and planning framework?

If yes, provide details of environmental policy and planning framework

6.4 Has the party taking the action previously referred an action under the EPBC Act, or been responsible for undertaking an action referred under the EPBC Act?

Page 36 of 44 Environment Protection and Biodiversity Conservation Act 1999

Provide name of proposal and EPBC reference number (if known)

Aerial baiting of Yellow Crazy Ant supercolonies;

(a) EPBC Referral 2002/722 (b) EPBC Referral 2009/5016 (c) EPBC Referral 2012/6438

Page 37 of 44 Environment Protection and Biodiversity Conservation Act 1999

7 Information sources and attachments (For the information provided above)

7.1 References  List the references used in preparing the referral.  Highlight documents that are available to the public, including web references if relevant.

Abbott, K.L. (2004). Alien ant invasion on Christmas Island, Indian Ocean: The role of ant-scale associations in the dynamics of supercolonies of the yellow crazy ant, Anoplolepis gracilipes. Ph.D Thesis, Monash University, Melbourne, Australia. http://arrow.monash.edu.au/vital/access/manager/Repository/monash:6496 (accessed 21 January 2013)

Abbott, K.L. (2005). Supercolonies of the invasive yellow crazy ant, Anoplolepis gracilipes, on an oceanic island: forager activity patterns, density and biomass. Insectes Sociaux 52: 266–273.

Abbott, K.L and Green, P.T (2007). Collapse of an ant-scale mutualism in a rainforest on Christmas Island. Oikos 116: 1238-1246.

Beeton, B., Burbidge, A., Grigg, G., How, R., McKenzie, N., and Woinarski, J. (2010). Final Report of the Christmas Island Expert Working Group to Minister for the Environment, Heritage and the Arts. Canberra, A.C.T., Australia. http://www.environment.gov.au/parks/publications/christmas/final- report.html (accessed 21 January 2013)

Boland, C.R.J., Smith, M.J., Maple, D.J., Tiernan, B., Barr, R., Reeves, R., and Napier, F. (2011). Heli-baiting using low concentration fipronil to control invasive yellow crazy ant supercolonies on Christmas Island, Indian Ocean. Pp. 152-156 in Veitch, C.R., Clout, M.N., and Towns, D.R. (eds.). Island invasives: eradication and management. Proceedings of the International Conference on Island Invasives. Gland, Switzerland: IUCN and Auckland, New Zealand. http://www.issg.org/pdf/publications/island_invasives/pdfhqprint/2boland.pdf (accessed 21 January 2013)

Butcher, R. and Hale, J. (2010). Ecological character description for The Dales Ramsar site. Unpublished report to the Department of the Environment, Water, Heritage and the Arts, Canberra, A.C.T., Australia. http://www.environment.gov.au/water/publications/environmental/wetlands/61-ecd.html (accessed 22 January 2013) Centre for Environment Stress and Adaptation (2011). Monitoring of the 2009 aerial baiting of yellow crazy ants (Anoplolepis gracilipes) on non-target invertebrate fauna on Christmas Island. Unpublished Report for the Director of National Parks, Canberra, A.C.T., Australia. http://www.environment.gov.au/parks/publications/christmas/fipronil-report.html (accessed 22 January 2013) Claussen, J. (2005). Native plants of Christmas Island. Flora of Australia, Supplementary Series 22. Australian Government Department of the Environment and Heritage, Australian Biological Resources Study. Canberra, A.C.T., Australia.

Commonwealth of Australia (2006). Threat Abatement Plan to Reduce the Impacts of Tramp Ants on Biodiversity in Australia and its Territories. Department of the Environment and Heritage, Canberra, A.C.T., Australia.

Page 38 of 44 Environment Protection and Biodiversity Conservation Act 1999

http://www.environment.gov.au/biodiversity/threatened/publications/tap/trampants.html (accessed 21 January 2013)

Davis, N.E., O’Dowd, D.J., Green, P.T., and Mac Nally, R. (2008). Effects of alien ant invasion on abundance, behaviour, and reproductive success of endemic island birds. Conservation Biology 22: 1165-1176. Davis, N.E., O'Dowd, D.J., Mac Nally, R., and Green, P.T. (2010). Invasive ants disrupt frugivory by endemic island birds. Biology Letters 6: 85-88.

Detto, T. and Tiernan, B. (2011). The 2011 island-wide survey report on Yellow Crazy Ants and Red Crabs. Unpublished report to Christmas Island National Park. Christmas Island.

Drew, M.M., Harzsch, S., Stensmyr, M., Erland, S., and Hansson, B.S. (2010). A review of the biology and ecology of the Robber Crab, Birgus latro (Linnaeus, 1767) (Anomura:Coenobitidae). Zoologischer Anzeiger 249: 45-67. Giraud, T., Pedersen, J.S., and Keller, L. (2002). Evolution of supercolonies: the Argentine ants of southern Europe. Proceedings of the National Academy of Science USA 99: 6075–6079.

Green, P.T. Comport, S., and Slip, D. (2004). The management and control of the invasive alien crazy ant (Anoplolepis gracilipes) on Christmas Island, Indian Ocean: the aerial baiting campaign, September 2002. Unpublished final report to Environment Australia and the Crazy Ant Steering Committee.

Green, P.T. and O’Dowd, D.J. (2009). Management of invasive invertebrates: lessons from the management of an invasive alien ant. Pp. 153-172 in Clout, M.N. and Williams, P.A. (eds). Invasive Species Management: A Handbook of Principles and Techniques. Oxford University Press, Oxford. Green, P.T., O’Dowd, D.J., Abbott, K.L., Jeffery, M., Retallick, K., and Mac Nally, R. (2011). Invasional meltdown: invader-invader mutualism facilitates a secondary invasion. Ecology 92: 1758-1768. Green, P.T., O’Dowd, D.J., and Lake, P.S. (1997) Control of seedling recruitment by land crabs in rain forest on a remote oceanic island. Ecology 78: 2474-2486. Green, P.T., O’Dowd, D.J., Lake, P.S. (2001). From resistance to meltdown: secondary invasion of an island rain forest. Pp. 451-455 in Ganeshaiah, K.N., Uma Shankar, R. and Bawa, K.S. (eds.). Tropical Ecosystems: Structure, Diversity and Human Welfare. Proceedings of the International Conference on Tropical Ecosystems, Bangalore, India. Oxford-IBH, New Delhi. Green, P.T., O’Dowd D.J., and Lake, P.S. (2008). Recruitment dynamics in a rainforest seedling community: context-independent impact of a keystone consumer. Oecologia 156: 373-385. Hale, J. and Butcher, R. (2010). Ecological Character Description for Hosnie’s Spring Ramsar Site. Unpublished report to the Department of the Environment, Water, Heritage and the Arts, Canberra, A.C.T., Australia. http://www.environment.gov.au/water/publications/environmental/wetlands/40-ecd.html (accessed 21 January 2013) Holmes, J. and Holmes, G. (2002). Conservation status of the flora of Christmas Island, Indian Ocean. Unpublished report to Environment Australia/Parks Australia North. Glenn Holmes and Associates, Atherton, Queensland. Kuhlmann, U., Schafner, U. and Mason, P.G. (2006). Selection of non-target species for host specificity testing. Pp. 15-37 in Bigler, F., Babendreier, D. and Kuhlmann, U. (eds.). Environmental

Page 39 of 44 Environment Protection and Biodiversity Conservation Act 1999

Impact of Invertebrates for Biological Control of Arthropods: Methods and Risk Assessment. CABI Publishing, Delemont, Switzerland. Lake, P.S. and O’Dowd, D.J. (1991). Red crabs in rainforest, Christmas Island: biotic resistance to invasion by an exotic snail. Oikos 62: 25-29.

Lowe, S. Browne, M., and Boudjelas, S. (2000). 100 of the world’s worst invasive alien species. Aliens 12: S1-S12. http://www.issg.org/database/species/search.asp?st=100ss (accessed 21 January 2013)

Lumsden, L. Schulz, M., Ashton, R., and Middleton, D. (2007). Investigation of the threats to the Christmas Island Pipistrelle. Unpublished report to the Department of Environment and Water Resources, Canberra, A.C.T., Australia. http::// http://ausbats.org.au/#/resources/4553704436 (accessed 21 January 2013)

Marr, R.M., O’Dowd, D.J., and Green, P.T (2003). Assessment of non-target impacts of Presto® 01 ant bait on litter invertebrates in Christmas Island National Park, Indian Ocean. Unpublished report to Parks Australia North, Darwin, NT. Neumann, G., Follett, P.A., Hollingsworth, R.G., and de León, J. (2010). High host specificity in Encarsia diaspidicola (Hymenoptera: Aphelinidae), a biological control candidate against the white peach scale in Hawaii. Biological Control 54: 107-113. Neumann, G., Green, P.T. and O’Dowd, D.J. (2011). First record of Pulvinaria urbicola (Hemiptera: Coccidae), a potentially damaging scale insect, on Christmas Island, Indian Ocean. Unpublished report to Parks Australia. http://www.environment.gov.au/parks/publications/christmas/pubs/pulvinaria-report-2011.docx (accessed 21 January 2013)

O’Dowd, D.J. and Green, P.T. (2000). Design and feasibility of an island-wide survey of the invasive alien ant Anoplolepis gracilipes and its impact on Christmas Island, Indian Ocean. Unpublished report to Parks Australia North, Darwin, NT.

O’Dowd, D.J. and Green, P.T. (2010). Invasional meltdown: do invasive ants facilitate secondary invasions? Pp. 271-272 in Lach, L., Parr, C. & Abbott, K. (eds.) Ant Ecology. Oxford University Press, Oxford. O’Dowd, D.J. Green, P.T., & Lake, P.S. (2003). Invasional ‘meltdown’ on an oceanic island. Ecology Letters 6: 812-817. http://wolfweb.unr.edu/~ldyer/classes/396/odowd.pdf (accessed 21 January 2013)

O’Dowd, D.J., Green, P.T., and Lake, P.S. (1999). Status, impact, and recommendations for research and management of exotic invasive ants in Christmas Island National Park. Unpublished report to Environment Australia, Canberra, A.C.T., Australia http://www.issg.org/database/species/reference_files/Christmas_Island_Report.pdf (accessed 21 January 2013).

O’Dowd, D.J., Green, P.T., Neumann, G., and Wittman, S. (2012). Executive summary: research and development for indirect biological control of the yellow crazy ant (Anoplolepis gracilipes) on Christmas Island, Indian Ocean. Unpublished report to the Director of National Parks and the Crazy Ant Scientific Advisory Panel, 27 pp. Palmer, W.A., Heard, T.A., and Sheppard, A.W. (2010). A review of Australian classical biological control of weeds programs and research activities over the past 12 years. Biological Control 52: 271-287. https://docs.google.com/a/monash.edu/viewer?a=v&q=cache:0-

Page 40 of 44 Environment Protection and Biodiversity Conservation Act 1999

K3mgvVy6oJ:northeast.landcarevic.net.au/oln/projects/soil-health-program/reference- documents/biocontrol-12-year- review/at_download/file+&hl=en&gl=au&pid=bl&srcid=ADGEESgGf1MZIbSjBgRLOJrC7PanstLsX FnzXToIee3sUdK7Dn6ilFKnFls9q89PVBlxOSUNgJKVCjUOAKMyOHG2WDsP2FkvkoWpsPpgoJ UntwXSJ6gKs63DvfDQ2aotz4z4J67pwN5K&sig=AHIEtbTK8Q9OLsVwz8aVroMgFYlNViLokA (accessed 21 January 2013) Schultz, M. (2004). National Recovery Plan for the Christmas Island Shrew Crocidura attenuata trichura. Commonwealth of Australia, Canberra. http://www.environment.gov.au/biodiversity/threatened/publications/recovery/c-attenuata- trichura/index.html (accessed 21 January 2013)

Schulz, M. and Barker, C. (2008). A Terrestrial Reptile Survey of Christmas Island, May-June 2008. Unpublished report for Parks Australia North, Darwin, NT.

Schulz, M. and Lumsden, L.F. (2004). National Recovery Plan for the Christmas Island Pipistrelle Pipistrellus murrayi. Commonwealth of Australia, Canberra. http://www.environment.gov.au/biodiversity/threatened/publications/recovery/p-murrayi/pubs/p- murrayi.pdf (accessed 21 January 2013) Secord, D. and Karieva, P. (1996). Perils and pitfalls in the host specificity paradigm. BioScience 46: 448-453. Smith, M.J., Boland, C.R.J., Maple, D.J., Scroggie, M., Tiernan B., and Napier. F. (unpublished results) The Christmas Island Red crab (Gecarcoidea natalis): temporal and spatial patterns in burrow counts. Smith. M.J., Cogger, H., Tiernan, B., Maple, D., Boland, C., Napier, F., Detto, T., and Smith, P. (2012). An oceanic island reptile community under threat: the decline of reptiles on Christmas Island, Indian Ocean. Herpetological Conservation and Biology 7: 206-218. Suhr, E.L., O’Dowd, D.J., Mackay, D.A., and McKechnie, S.W. (2011). Genetic structure, behaviour and invasion history of the Argentine ant in Australia. Evolutionary Applications 4: 471–484. Sunamura, E., Espadaler, X., Sakamoto, H., Suzuki, S., Terayama, M. and Tatsuki, S. (2009). Intercontinental union of Argentine ants: behavioral relationships among introduced populations in Europe, North America, and Asia. Insectes Sociaux 56: 143–147. Thomas, M.L., Becker, K., Abbott, K., and Feldhaar, H. (2010). Supercolony mosaics: two different invasions by the yellow crazy ant, Anoplolepis gracilipes, on Christmas Island, Indian Ocean. Biological Invasions 12: 677-687. Wetterer, J.K. (2005). Worldwide distribution and potential spread of the long-legged ant, Anoplolepis gracilipes (Hymenoptera: Formicidae). Sociobiology 45: 77-97.

7.2 Reliability and date of information For information in section 3 specify:  source of the information;  how recent the information is;  how the reliability of the information was tested; and  any uncertainties in the information.

Information sources: Full references are provided in Section 7.1 for the cited sources in Section 3. In addition, other information came from the Draft Christmas Island National Park Management Plan

Page 41 of 44 Environment Protection and Biodiversity Conservation Act 1999

(Commonwealth of Australia 2012), the Environment Protection and Biodiversity Conservation Act 1999 (Cwth), and the List of Commonwealth National Heritage places. Information age: The information varies in age from 1991 through to 2012. Information reliability: Studies published in scientific journals have been through a process of peer review and thus can be considered reliable. Unpublished reports have not been through this process, but can be considered as reliable because their authors are recognized as knowledgeable experts. This referral draws heavily on an unpublished report to the Director of Parks and the Crazy Ant Scientific Advisory Panel (O’Dowd et al. 2012). This report contains unpublished research, and in order to protect the publication rights of its authors this research should not be available to the public alongside the referral. The research was, however, formally peer-reviewed and endorsed by CASAP on 12 December 2012 (Attachment 3). Information uncertainties: Uncertainties about the implementation and outcomes of the proposed action have been detailed in sections 2.1, 2.2, 2.4 and 2.5.

7.3 Attachments Indicate the documents you have attached. All attachments must be less than two megabytes (2mb) so they can be published on the Department’s website. Attachments larger than two megabytes (2mb) may delay the processing of your referral.

 attached Title of attachment(s) You must attach figures, maps or aerial photographs Included within the showing the project locality (section 1) referral document figures, maps or aerial photographs showing the location of the project in Included within the respect to any matters of national referral environmental significance or important document features of the environments (section 3) If relevant, attach copies of any state or local government approvals and consent conditions (section 2.5) copies of any completed assessments to meet state or local government approvals and outcomes of public consultations, if available (section 2.6) copies of any flora and fauna investigations and surveys (section 3) technical reports relevant to the assessment of impacts on protected matters that support the arguments and conclusions in the referral (section 3 and 4) report(s) on any public consultations undertaken, including with Indigenous stakeholders (section 3)

Page 42 of 44 Environment Protection and Biodiversity Conservation Act 1999

Other attachments Attachment 1. Indirect  biological control of the yellow crazy ant on Christmas Island – host range testing of the proposed agent Tachardiaephagus somervillei for the primary target Tachardina aurantiaca.

Attachment 2. Biological Control Implementation for Tachardina aurantiaca on Christmas Island.

Attachment 3. CASAP advice to DNP on the feasibility of indirect biocontrol for the suppression of YCA supercolonies

Attachment 4. Nomination of the Yellow lac scale as a biological control agent target

Page 43 of 44 Environment Protection and Biodiversity Conservation Act 1999

8 Contacts, signatures and declarations NOTE: Providing false or misleading information is an offence punishable on conviction by imprisonment and fine (s 489, EPBC Act).

Under the EPBC Act a referral can only be made by:  the person proposing to take the action (which can include a person acting on their behalf); or  a Commonwealth, state or territory government, or agency that is aware of a proposal by a person to take an action, and that has administrative responsibilities relating to the action1.

Project title: Importation, rearing and release of Tachardiaephagus somervillei (Hymenoptera: Encyrtidae) as a biological control agent for the yellow lac scale Tachardina aurantiaca (Hemiptera: Kerriidae) on Christmas Island, Indian Ocean 8.1 Person proposing to take action This is the individual, government agency or company that will be principally responsible for, or who will carry out, the proposed action.

If the proposed action will be taken under a contract or other arrangement, this is:  the person for whose benefit the action will be taken; or  the person who procured the contract or other arrangement and who will have principal control and responsibility for the taking of the proposed action.

If the proposed action requires a permit under the Great Barrier Reef Marine Park Act2, this is the person requiring the grant of a GBRMP permission.

The Minister may also request relevant additional information from this person.

If further assessment and approval for the action is required, any approval which may be granted will be issued to the person proposing to take the action. This person will be responsible for complying with any conditions attached to the approval.

If the Minister decides that further assessment and approval is required, the Minister must designate a person as a proponent of the action. The proponent is responsible for meeting the requirements of the EPBC Act during the assessment process. The proponent will generally be the person proposing to take the action3. Name Peter Cochrane Title Commonwealth Director of National Parks. Organisation Director of National Parks (Christmas Island National Park) ACN / ABN (if applicable) 13 051 694 963 Postal address GPO Box 787 Canberra Act 2601

Telephone (02) 6274 2220 Email [email protected] Declaration I declare that to the best of my knowledge the information I have given on, or attached to this form is complete, current and correct. I understand that giving false or misleading information is a serious offence.

1 If the proposed action is to be taken by a Commonwealth, state or territory government or agency, section 8.1 of this form should be completed. However, if the government or agency is aware of, and has administrative responsibilities relating to, a proposed action that is to be taken by another person which has not otherwise been referred, please contact the Referrals Business Entry Point (1800 803 772) to obtain an alternative contacts, signatures and declarations page.

2 If your referred action, or a component of it, is to be taken in the Great Barrier Reef Marine Park the Minister is required to provide a copy of your referral to the Great Barrier Reef Marine Park Authority (GBRMPA) (see section 73A, EPBC Act). For information about how the GBRMPA may use your information, see http://www.gbrmpa.gov.au/privacy/privacy_notice_for_permits.

3 If a person other than the person proposing to take action is to be nominated as the proponent, please contact the Referrals Business Entry Point (1800 803 772) to obtain an alternative contacts, signatures and declarations page.

Page 44 of 44 Environment Protection and Biodiversity Conservation Act 1999

I agree to be the proponent for this action. I acknowledge that I may be liable for fees related to my proposed action following the introduction of cost recovery under the EPBC Act.

Signature Date 15 April 2013

8.2 Person preparing the referral information (if different from 8.1) Individual or organisation who has prepared the information contained in this referral form. 1* - Drs Peter Green and Dennis O’Dowd Name 2* - Michael Misso 1* - Senior Lecturer / Reader (retired) Title 2* - Manager, Christmas Island National Park 1* - La Trobe University Organisation 2* - Parks Australia 1* - 64 804 735 113 ACN / ABN (if applicable) 2* - 13 051 694 963 1* - Department of Botany, Biological Sciences Building 2 Postal address 2* - Christmas Island National Park PO Box 867, Christmas Island, Indian Ocean. 1* - 03 9479 3675 Telephone 2* - 08 9168 8700 1* - [email protected] Email 2* - [email protected] Declaration I declare that to the best of my knowledge the information I have given on, or attached to this form is complete, current and correct. I understand that giving false or misleading information is a serious offence. 8/4/2013

Signature Date

Page 45 of 44 Environment Protection and Biodiversity Conservation Act 1999

REFERRAL CHECKLIST NOTE: This checklist is to help ensure that all the relevant referral information has been provided. It is not a part of the referral form and does not need to be sent to the Department.

HAVE YOU:  Completed all required sections of the referral form?  Included accurate coordinates (to allow the location of the proposed action to be mapped)?  Provided a map showing the location and approximate boundaries of the project area?  Provided a map/plan showing the location of the action in relation to any matters of NES?  Provided complete contact details and signed the form?  Provided copies of any documents referenced in the referral form?  Ensured that all attachments are less than two megabytes (2mb)?  Sent the referral to the Department (electronic and hard copy preferred)?

Page 46 of 44 Attachment 1

Host Range Testing of the Biological Control Agent for the Primary Target (Tachardina aurantiaca)

Background

No endemic scales were found on Christmas Island (CI) during the search for possible non- target organisms. All scale insects which could be considered “non-target” organisms are exotic. However, while we are highly confident that there are no non-target organisms on CI that could be harmed by the importation of a biological control agent (BCA) for Tachardina.aurantiaca (TA) and no direct or indirect non-target effects are expected (at a high level of certainty), the non-existence of an endemic non-target species cannot be proven by means of empirical science (a certain probability of the existence of a non-target remains however low). Therefore, a BCA with a narrow host range is preferred to a generalist BCA. As no non-targets from CI can be tested, the principle of “centrifugal host range testing” will be used to assess the host range of the BCA (Kuhlmann et al 2006, Neumann et al. 2010). This means that species other than the known host species of the BCA are tested with the most closely related species (least phylogenetically distant, close similarities in biology and ecology) tested first then less similar species (i.e., the“centrifugal principle”).

The main factors affecting (limiting) host ranges of parasitoids

1) Phylogenetic distance (from close relationship to more distant relationship – usually based on taxonomical relatedness) 2) Host ecology

Predicting relative host range based on parasitoid biology

Godfray (1994) suggested that the relative host ranges of parasitoids can be predicted based on the biology of the parasitoid (biochemical and physiological connection with hosts). The more “intimate” the physiological relationship, probably the narrower the host range:

- koinobiont parasitoids (those that allow their host to continue development for some time while feeding on the host before finally killing it) that are endoparasitic (develop inside the body of the host) probably have fewer hosts as they have to overcome active host defenses and immune responses - idiobiont parasitoids (those that prevent their host to continue development after the attack), which are also ectoparasitic (develop outside of the hosts body) probably have more hosts as they don’t have to overcome the hosts defenses - koinobiont parasitoids attacking phylogenetically isolated hosts probably have a narrow host range

General considerations when selecting test species to determine host range

- manageable number of species (10 to 15 but less are acceptable if high specificity is found initially) - use species that are both similar to known host taxonomically and ecologically - test more than one species in categories (e.g. more than one species that live in the same habitat and are related to known host) - use an ‘out-group’ – a species that is related to known host to some degree but farther than the initial species – use the out-group early in the process - availability - species that can be lab-reared are preferred but field test are acceptable if controlled

Testing

- no-choice tests are the most stringent - choice tests can follow no-choice tests if tested species is accepted as host - behavioral observations are critical (host feeding, probing – can lead to non-target mortality even if non-target is not accepted as a host)

Host range testing for Tachardiaephagus somervillei

The first step in assessing the host range of T. somervillei is to study its biology so some predictions of the breadth of its host range (see above). This research is ongoing in Kuching, Sarawak, Malaysian Borneo and at the Forest Research Institute of Malaysia in Selangor, West Malaysia and results are expected by early 2013. It must be emphasized that predictions based on the biology and behavior of the parasitoid can be of limited value. Nonetheless, such information will be helpful

The second step in assessing the host specificity of the BCA will be locating related potential hosts (scale insects) in the same habitat in which the known host (TA) lives. Ideally, other scale insects feeding on the same host plant on which TA is present and T. somervillei parasitism is present will be located. If no parasitism of this potential host is observed in field conditions, no-choice test will be carried out in the laboratory. The first preference will be given to members of Coccidae and other neococcoid taxa occupying the same habitat as TA. An out-group from archeococcid taxa will be selected early in the process as well. The exact list of species will be based on availability and the phylogenetic centrifugal sequence of testing these species will be determined with the advice of Dr. Penny Gullan (ANU), a specialist on the phylogeny of scale insects.

Testing approaches

1) Field surveys. These surveys in the area where the BCA can be found provide information about how the BCA utilizes available, potential hosts. The shortcoming of this approach is that it is essentially a “choice test” meaning that a potential host may not be utilized simply because a more preferred host is available. 2) “No-choice tests” including behavioral observations. In these tests, the BCA is presented with the test species without any alternative which maximizes oviposition pressure. The results of these tests must be evaluated on different levels: a) the potential host can serve as a host and BCA progeny is produced (suitable host), b) oviposition occurs but progeny is not produced (host not suitable but attacked resulting in mortality of the test species), c) no oviposition but probing or probing/host feeding occurs resulting in mortality of the test species, and d) test species rejected, no oviposition, no probing or probing/host feeding. The no-choice tests are very stringent and any test resulting in “d” outcome should determine the test species non-susceptible at any level. All no-choice tests must include a control group where BCA individuals used in the test are randomly selected and provided with known host (TA) to ensure that the quality of the BCA used in the test is good and the experimental conditions do not negatively influence parasitism. 3) “Choice tests”. The no-choice tests present the BCA with an option of not being able to produce offspring which can result in such high oviposition pressure that the test species is attacked while in natural conditions this would not happen. If a test species is attacked in no-choice tests (but is not a suitable host, that is, progeny is not produced but mortality occurs), choice tests will be conducted where the test species and the known host (TA) will be presented to the BCA simultaneously. If in this test the test species is not attacked consistently or at a very low rate, the test species should be considered as susceptible but unsuitable and should not be considered as a host.

We propose that if no test species (note that the test species list is yet to be determined based on availability and phylogenetic considerations) proves susceptible at any level the expected non-target effect should be considered highly insignificant. If some of the test species are susceptible but unsuitable, coupled with the very low probability of non-targets present on Christmas Island, the expected non-target effect should be considered insignificant. If some test species prove suitable hosts and are attacked consistently in choice tests, the introduction of the BCA to Christmas Island should be reconsidered meaning considering an alternative BCA or weighing the risk of introducing a BCA with a wide host range but with most likely no non-targets on Christmas Island versus significant ecosystem loss if no BCA is introduced.

References cited

Godfray, H. C. J. 1994. Parasitoids: Behavioral and Evolutionary Ecology. Princeton University Press, Princeton, New Jersey, pp. 447.

Kuhlmann, U., U. Schafner and P. G. Mason. 2006. Selection of non-target species for host specificity testing. In: Environmental Impact of Invertebrates for Biological Control of Arthropods: Methods and Risk Assessment (eds. F. Bigler et al.). CAB International.

Neumann, G., P. A. Follett, R. G. Hollingsworth, and J. de León. 2010. High host specificity in Encarsia diaspidicola (Hymenoptera: Aphelinidae), a biological control candidate against the white peach scale in Hawaii. Biol. Control. 54(2): 107-113. 1

Attachment 2

Biological Control Implementation for Tachardina aurantiaca on Christmas Island

Gabor Neumann, Peter T. Green, Dennis J. O’Dowd

GOAL

Import, rear, release, and monitor the efficacy of an approved biological control agent on Christmas Island to control the yellow lac scale Tachardina aurantiaca for the long-term management of the introduced yellow crazy ant (Anoplolepis gracilipes).

OBJECTIVES

1. Develop and implement mass-rearing efforts for approved and permitted biological control agent (BCA) of Tachardina aurantiaca (TA hereafter).

2. Release of permitted biological control agent in areas infested with TA where densities of the yellow crazy ant (YCA hereafter) are at supercolony levels or where the potential of supercolony formation exists.

3. Monitor survival, establishment of released biological control agent, and its impact on TA populations and YCA densities.

4. Adapt methods to improve the rearing, release and monitoring of TA biological control agent.

PERFORMANCE CRITERIA:

1. Successful importation of the BCA into a rearing facility on the island. 2. Successful mass-rearing of the BCA in glasshouse/screenhouse facilities. 3. Release and establishment of BCA at selected field sites. 4. High mortality and reduced densities of TA at field sites. 5. Spread and establishment of the BCA at new sites.

ACTIONS

1. Develop and implement mass-rearing efforts

1.1. Maintain laboratory research colonies of TA biological control agent in Christmas Island National Park (CINP) facilities to develop mass-rearing procedures and, if needed, continue research efforts into natural enemy biology and behavior. Establish a separate colony in CINP facilities that will serve as a main production colony and also as back up at a separate location. It is highly desirable that specimens of the biological control agent be collected periodically in the area of origin, and after 2

phytosanitary procedures to insure removal of any pathogens and hyperparasitoids, incorporated into the Christmas Island colonies in order to minimize the selection of laboratory-adapted insects.

1.2. The primary role of the production facility is to mass-rear biological control agents for field release. This includes the production of optimal host life stages of TA to support rearing of the biological control agent. Consequently, host plants for TA must be produced as well. TA and its natural enemies may be difficult to mass-rear due to the relatively long life cycle of TA and the need for fresh host plants (that may or may not be reused) at a regular basis. Depending on the difficulty of rearing the biological control agent, their availability for releases maybe limited at any given time. As mass-rearing methods improve and production increases, the goal will be to be able to provide the biological control agent for releases in all areas as needed.

Requirements:

1. Host plant production. Small host plants (e.g., Inocarpus seedlings) for rearing TA can be produced at the existing nursery. Host plant production levels will depend largely on two factors - the available space and the labour that CINP can commit to producing host plants. 2. Facility for mass rearing the biological control agent. The major infrastructure item for this project would be a dedicated glasshouse/screenhouse (preferably closed, but solid-roof screen house would probably be sufficient) for rearing of the biological control agent. Suitable host plants will be transferred to the glasshouse/screenhouse. If chemical control of pests and diseases on host plants is needed at the nursery, host plants must remain uninfested in the glasshouse/screenhouse until insecticidal residues are no longer active. Host plants can be inoculated with TA within the same facility as long as plants in "waiting" and infested plants are separated with a barrier that prevents TA crawler movement (e.g., a water moat system). Availability of a second, smaller glasshouse/screenhouse, to serve as a backup should the colony of natural enemies fail in the primary glasshouse, would be optimal, although not absolutely necessary. Alternatively, a laboratory facility, similar in size to the present biological control laboratory (i.e. the demountable) at CINP, should be sufficient if all space was dedicated to production. However, this would require sufficient lighting and temperature control to sustain host plants during rearing of biological control agents on their host, TA. 3. Research and office space. Sufficient laboratory and office space is needed to support a biocontrol professional and two technical staff (see Personnel below). The existing laboratory space or equivalent currently dedicated to the biological control program at CINP should be sufficient. 4. Personnel. One full-time biological control professional (BCP) is needed for the production research and mass-rearing over the three-year project. As part of the rehabilitation program, the nursery personnel currently produce host plants, but given the time needed for host plant production and the competing needs of the rehab nursery, dedicated support will be necessary. One part-time position would probably sufficient. As back up, it is advisable that two CINP personnel are trained in the basic maintenance of the mass-rearing colonies.

3

2. Release of permitted biological control agent

2.1 Establish criteria for choosing suitable primary release sites that could include the following: (a) positive evidence of TA infestation; (b) relatively high percentage of TA host plants in the overstorey; (c) predominance or high occurrence of TA host plants in the understorey; and, (d) presence of YCA supercolony or at least YCA presence; and, (e) site not subjected to current pesticide exposure (contact or systemic) and that, if possible, are free from any pesticide residues.

These criteria for site selection for release could be gleaned from the biennial IWS followed by site inspections. Each release will consist of a specified number of the suitable biological control agent. The number released would be determined based on prior knowledge about the biological control agent, the level of TA infestation, size of target area, and the capacity of the rearing facility. The releases must occur when the appropriate TA life stage(s) is (are) present at a site. Depending on agent availability, timing of releases will be different for different locations.

2.2. Develop a TA Biological Control Manual and provide training to Christmas Island National Park personnel to aid in understanding and conducting the choosing of suitable primary release sites and release of the biological control agent.

2.3. Initiate releases of TA biological control agent at primary release sites. Release methods will largely depend on knowledge currently being gathered about the biological control agent in Malaysia.

Requirements:

Some additions to the IWS, involving increased work load for the field crew, would be required to facilitate release site selection. Releases would be conducted as part of the duties of the biocontrol professional and the CINP support staff. The TA Biological Control Manual would be produced by the biocontrol professional in discussion with CINP; the biocontrol professional would lead training of CINP personnel. The CINP field crew has already received training in identifying and collecting scale insects and parasitoids for the BCP.

3. Determine establishment and monitor spread and impacts

3.1. Develop detection protocols and train CINP personnel to be able to easily detect presence of the biocontrol agent. This will be important not only during establishment monitoring at release sites but also to detect possible dispersal of the biological control agent to areas where it was not released. It may be feasible to use the biennial IWS to document any spread, at least onto understorey seedlings and saplings, or TA infested sentinel plants placed in the field at the release and control sites.

3.2. Develop protocols to estimate biological control agent impact on TA populations and YCA densities, considering pre-post release monitoring and comparisons of release and control sites. This will include parasitism rates, TA progeny production and survival, biological control agent dispersal, and YCA density changes over time. Dedicate control areas where TA and YCA are present but the biological control is not released. The number of control sites will be determined based on the release sites and area availability. CINP will determine how many of 4

the selected primary release sites will be actually available for release and the available control areas where no YCA management practices (i.e., application of toxic ant baits) will be applied. Ideally, control sites should be distant enough from release sites so that the chances for biological control agent dispersal are low for a reasonable period of time.

Requirements:

If the IWS is used to coarsely monitor spread and impact of the biological control agent, additional workload would be required for the field crew in the next and subsequent IWS. The biological control professional and technical support staff should be sufficient to monitor more details about establishment, spread, and parasitization rates at release sites.

4. Adapt methods

Continue improving mass-rearing procedures to optimize production efficiency. Improve release procedures based on first releases to ensure efficiency of release procedures, biological control agent survival after release and establishment at release sites.

Requirements:

The biocontrol professional will lead methods development with the advice of CINP. No additional facilities should be needed for continued development of methods and protocols; the facilities established during research and production will suit the needs of the release, establishment, and impact monitoring.

5. Staffing

Biological control professional. The BCP is critical for the full 3-year period of the project and would be active in all aspects of the program (Actions 1-4). The BCP would lead development of mass- rearing techniques and quality assurance of the BCA, develop release and monitoring protocols in conjunction with CINP, and adapt procedures if and when circumstances demand. The BCP will develop a Biological Control Manual to cover all aspects of mass rearing, including quality assurance, release site selection, and a monitoring program evaluating the efficacy of the BCA. Training will also be provided to CINP personnel to assist in understanding and conducting the choosing of suitable primary release sites, release of the BCA, and monitoring of its efficacy in controlling TA.

Technical support. Skilled technical support will be needed that are dedicated to the project and under the supervision of Biological Control Professional. These could be CINP staff assigned to different aspects of the project.

6. Timeline

Although it is difficult at present to forecast a firm timeline for the program, several key program objectives have already been accomplished. (1) TA has been located over a 1900-km range in Malaysia. Specimens from sites in Peninsular Malaysia and Malayan Borneo match those on Christmas Island morphologically and genetically. TA in Malaysia is rare and patchy, consistent with control by natural enemies. (2) The assemblage of natural enemies of TA has been determined for a 5

number of sites in Malaysia and six primary parasitoid species have been identified. Parasitism rates are high at all sites examined so far. (3) A prospective BCA, the parasitoid Tachardiaephagus somervillei (Hymenoptera: Encyrtidae) has been identified and research is underway to understand its biology. (4) Techniques for rearing host plants, the lac scale TA, and the prospective BCA have been developed at FRIM in Kuala Lumpur that should be applicable to development of mass-rearing on Christmas Island. The greatest uncertainty is the timeline required for permitting the import and release of the biological control agent, especially since an approvals process for the BCA has not yet been identified for the external territory of Christmas Island. However, given that appropriate natural enemies are available in SE Asia, we believe it should be possible to complete biological control agent candidate selection, non-target effect research (host specificity of the BCA), and the preparation of an environmental impact assessment by the end of 2013 and perhaps earlier if an approvals process can identified.

From the point of importation of the BCA, the following time requirement is estimated:

1. Development and implementation of mass-rearing: 6 months - 1 year (depending on initial success). 2. Release, establishment/impact monitoring, and continued methods development: 3 years. The biological control agent should be released in three, consecutive dry seasons with concomitant monitoring of establishment and impact on TA and YCA. The need for further releases would be determined after the initial three-year releases.

Assuming that the importation of the biological control agent(s) would be permitted in late 2013, mass rearing could start in late 2013. If natural enemy production is successful, the first release year would be 2014 and continue in 2015 and 2016. Pre-release monitoring of scale insects and associated YCA levels can be conducted in 2013 and it may be possible to include some of this monitoring in the IWS. Post-release monitoring should be conducted regularly during 2014, 2015 (complemented by the island-wide survey in 2015), and 2016. The last releases should be done in 2016. Additional monitoring should be conducted (along with the island-wide survey) in 2017 to determine whether further releases are necessary. If natural enemy establishment is evident and scale insect/YCA control is promising, mass-rearing will be no longer necessary but maintaining the biological control agent in the laboratory at smaller numbers is advisable in case further releases are necessary later.

CSIRO ECOSYSTEMS SCIENCES

Tropical Ecosystems Research Centre PMB 44 Winnellie, NT 0822, Australia T (03) 8944 8400 • ABN 41 687 119 230

15 January 2013

Mr Mike Misso Manager, Christmas Island National Park [email protected]

Dear Mike,

During CASAP’s meeting on 12 December 2012 we considered the executive summary of the report ‘Research and development for indirect biological control of the yellow crazy ant (Anoplolepis gracilipes) on Christmas Island, Indian Ocean’. Drs Green and O’Dowd were excused from this consideration due to their significant roles in the writing of the report.

The report summarises results of research related to the potential use of biocontrol of the scale insect Tachardina aurantiaca as a means of controlling crazy ant infestations on Christmas Island. The research shows that: (1) The supply of liquid carbohydrate plays a key role in the dynamics of crazy ant colonies on CI; (2) T. aurantiaca is a key provider of liquid carbohydrate on CI; (3) In contrast to CI, within its native range in Malaysia T. aurantiaca has a diverse range of natural enemies, and is rare and patchily distributed; (4) In Malaysia, T. aurantiaca has high rates of parasitism by the encyrtid Tachardiaephagus somervillei , which appears to specialise on the family Kerriidae (not native to CI) and can be reared under laboratory conditions. The report concludes that prospects of successful biocontrol are good, and therefore the biocontrol program should proceed.

CASAP members were unanimously of the view that the research had very high scientific merit, and effectively demonstrates the potential for successful biocontrol. It was considered that the risks associated with introducing a bio-control agent were low, particularly given existing bio-security risks on CI. CASAP endorses the conclusion of the report that the biocontrol program should proceed, and recommends that it covers soft scale as well as T. aurantiaca, since it is highly feasible and requires minimal extra resources to do so.

Yours sincerely,

Prof Alan N. Andersen Chair, Christmas Island Crazy Ant Scientific Advisory Panel [email protected] (08) 8944 8431

Attachment 4

Nomination of the Yellow lac scale as a biological control agent target

Target: Yellow lac scale Tachardina aurantica Cockerell

Nominating Organisation:

Director of National Parks

Prepared by Gabor Neumann, Peter Green, and Dennis O’Dowd LaTrobe University Bundorra, Vic 3086

1 1. Taxonomy Class: Insecta Order: Hemiptera Sub-order: Sternorrhyncha Superfamily: Coccoidea Family: Kerriidae Subfamily: Tachardininae Genus: Tachardina Species: aurantiaca Cockerell

Common names: Yellow lac scale, Golden lac scale, Mooncake lac scale

Synonyms: Not applicable

2. Description Female scales usually separate, sometimes coalescing, round when seen from above, 4 mm. long, convex, but flattened dorsally, so that they are not half as high as broad; surface thrown more or less into concrete folds; colour bright yellow; median dorsal area ferruginous, with radiating ridges and the usual orifices, the minutely transversely ribbed larval exuviae in the middle. Young up to about 2 mm long, orange-ferruginous, with rather obscure radiating ridges. Resinous test of adult female (Fig. 1) circular, somewhat flattened dorsally: the larval pellicle forming a crenulated ridge in the centre of the dorsal area: anal orifice circular or broadly oval, its posterior rim raised into a prominent tooth- like point: respiratory orifices small, very slightly prominent, situated one on each side of and close to the larval pellicle: sides more or less distinctly broadly radially fluted. Colour bright fulvous to castaneous, the larval pellicle reddish; semitranslucent (Ben-Dov 2012). The test (cover) of males is elongate and red or reddish brown with an operculum at the posterior end through which the winged, red males emerge. Males are short-lived and may not play an important role in reproduction (Fig. 2).

Fig. 1. Mature females of the yellow lac Fig. 2. Male tests of the yellow lac scale scale tended by the yellow crazy ant and an emerged male. Photo: G.Neumann Anoplolepis gracilipes. Photo: S. Belcher

2 3. Distribution

3.1 Native geographic range and centre of origin Southeast Asia: Indonesia (Java); Malaysia (Peninsular Malaysia, Sarawak, Sabah); Singapore; Thailand (Ben-Dov 2012). Likely centre of origin is Sundaland in Southeast Asia.

3.2 Australian and overseas distribution Indian Ocean islands: Christmas Island, Indian Ocean (External territory of Australia) (Campbell 1964; Ben-Dov 2012); Maldives, Indian Ocean (Ben-Dov 2012) 4. Proposing Organisation Director of National Parks 5. Pest status and costs The yellow lac scale is an invasive environmental pest on Christmas Island (Indian Ocean). A broad host plant generalist, it attacks at least 29 native plant species on the island (Abbott 2004). Its mutualistic association with the exotic yellow crazy ant Anoplolepis gracilipes leads to population build-up of both species and large impacts on the rainforest ecosystem - so-called invasional meltdown (O’Dowd et al. 1999; O’Dowd et al. 2003). Direct effects include the rapid decline of the dominant rainforest tree Inocarpus fagifer and widespread forest dieback. Indirect impacts include the population collapse of the dominant keystone species (the red land crab) on the island (O’Dowd et al. 2003), negative effects on endemic land birds (Davis et al. 2008, 2010), and invasion and rapid spread of the giant African land snail Achatina fulica in island rainforest (Green et al. 2011). Invasion by yellow crazy ants on Christmas Island has been Listed since 2005 as a Key Threatening Process under the EPBC Act 1999 (DSEWPaC 2005). Furthermore, the association between honeydew-secreting scale insects and invasive ants was recognized as a key threat to biodiversity on Christmas Island (Christmas Island Expert Working Group 2010), and the Australia government accepted the recommendation of the Expert Working Group to accelerate biological control efforts for scale insects on Christmas Island (Commonwealth of Australia 2011).

Current annual recurrent costs of mitigating the damage caused by the association of scale insects with the yellow crazy ant are estimated at up to $1 m dollars a year.

6. Beneficial status The yellow lac scale is a major environmental pest in its area of introduction and invasion on the Australian External Territory of Christmas Island, and is not known as a beneficial insect in its centre of origin.

3 7. Control methods

7.1 Chemical control There are no known direct chemical control methods for the yellow lac scale. Population outbreaks are suppressed indirectly when its tending ant mutualist, the yellow crazy ant, is controlled with toxic bait (Abbott and Green 2007, Green and O’Dowd 2010). However, yellow lac scale and crazy ant populations frequently recover in baited areas and form new infestations elsewhere, necessitating repeated chemical control for more than a decade. Moreover the toxic bait containing fipronil has some undesirable non-target impacts.

7.2 Biological control On Christmas Island, the yellow lac scale is attacked by one hymenopteran associate and two species of scale predatory Lepidoptera (Table 1). Marietta leopardina exclusively attacks male yellow lac scales, but given the continued high densities of lac scales in supercolonies of invasive crazy ants, this parasitoid is clearly not an effective biological control agent. The two lepidopteran predators (Eublemma sp. and ?Holcocera sp.) are rare and do not seem to have any significant effect on yellow lac scale populations on the island.

Table 1. Insect biological control agents of the yellow lac scale Tachardina aurantiaca on the Australian external territory of Christmas Island.

Order and family Species

Hymenoptera: Encyrtidae Marietta leopardina Motschulsky

Lepidoptera: Noctuidae Eublemma sp. (tentative) Blastobasidae Holcocera sp. (tentative)

8. Potential Conflicts of Interest There are no known conflicts of interest associated with the proposed biological control of the yellow lac scale. Its only known impacts on Christmas Island are negative, and it is not an important food item for any native species. Its association with the exotic yellow crazy ant constitutes a threat to the biodiversity and conservation values of the Christmas Island National Park (Christmas Island Expert Working Group 2010, Commonwealth of Australia 2011). Furthermore, the cost of mitigating the damage caused by the association between the yellow lac scale and the yellow crazy ant has cost the Australian government millions of dollars since baiting commenced in 1999.

References

Abbott, K. (2004) Alien ant invasion on Christmas Island, Indian Ocean: the role of ant- scale associations in the dynamics of supercolonies of the yellow crazy ant Anoplolepis gracilipes. PhD thesis, Monash University, Melbourne, Victoria, Australia.

Abbott, K. and Green, P.T. (2007). Collapse of ant-scale mutualism in a rainforest on Christmas Island. Oikos 116, 1238-1246.

4 Ben-Dov, Y. 2012. ScaleNet, Tachardina aurantiaca. 13 June 2012. http://www.sel.barc.usda.gov/catalogs/kerriida/Tachardinaaurantiaca.htm

Christmas Island Expert Working Group (2010). Final report of the Christmas Island Expert Working Group to the Minister for Environment, Heritage and the Arts. http://www.environment.gov.au/parks/publications/christmas/pubs/final-report.pdf

Commonwealth of Australia (2011). Australian government response to the recommendations of the Christmas Island Expert Working Group. http://www.environment.gov.au/parks/publications/christmas/pubs/ewg-report- response.pdf

Davis, N.E., O’Dowd, D.J., Green, P.T., and Mac Nally R.T. (2008) Effects of an alien ant invasion on abundance, behavior, and reproductive success of endemic island birds. Conservation Biology 22, 1165-1176.

Davis, N.E., O’Dowd, D.J., Mac Nally R.T., and Green, P.T. (2010) Invasive ants disrupt frugivory by endemic island birds. Biology Letters 6, 85-88.

DSEWPaC (2005). Loss of biodiversity and ecosystem integrity following invasion by the Yellow Crazy Ant (Anoplolepis gracilipes) on Christmas Island, Indian Ocean. http://www.environment.gov.au/cgi-bin/sprat/public/publicshowkeythreat.pl?id=16

Campbell, T.G. 1964. Entomological survey of Christmas Island (Indian Ocean) with special reference to the insects of medical, veterinary, agricultural and forestry significance.

Green, P.T. and O’Dowd, D.J. (2010). Management of invasive invertebrates: lessons from the management of an invasive alien ant. Pp. 153-172 in (Clout, M. and P. Williams, eds.) Invasive species management: A Handbook of Principles and Techniques. Oxford University Press, Oxford.

Green, P.T., O’Dowd, D.J., Abbott, K., Jeffrey, M., Retallick, K., and Mac Nally, R.T. (2011). Invasional meltdown: Invader–invader mutualism facilitates a secondary invasion. Ecology 92, 1758-1768.

O’Dowd, D.J., Green, P.T., and Lake, P.S. (1999) Status, impact, and recommendations for research and management of exotic invasive ants in Christmas Island National Park. Report to Environment Australia. http://www.issg.org/database/species/reference_files/Christmas_Island_Report.pdf

O’Dowd, D.J., Green, P.T., and Lake, P.S. (2003). Invasional ‘meltdown’ on an oceanic island. Ecology Letters 6, 812-817.

5 Amendments to Referral number 2013/6836

Importation, rearing and release of Tachardiaephagus somervillei (Hymenoptera: Encyrtidae) as a biological control agent for the yellow lac scale Tachardina aurantiaca (Hemiptera: Kerriidae) on Christmas Island, Indian Ocean

Issue 1 - Regulatory Framework In several places in the submitted referral document, uncertainty was expressed about the appropriate regulatory framework under which the proposal to import and release a biological control agent on Christmas Island would be considered. These were sections • 1.11 • 2.1, subsection entitled Approval, Importation & Rearing, and Release & Monitoring of Tachardiaephagus • 2.4, subsection entitled Regulatory Framework – Commonwealth or state legislation or policies under which approvals are required or will be considered against • 2.5 subsection entitled Environmental impact assessments under Commonwealth, state or territory legislation • 4, entitled Measures to avoid or reduce impacts Resolution: Christmas Island falls under the Quarantine Act 1908 , and subordinate legislation in the Quarantine (Christmas Island) Proclamation 2004 . Under this legislation, the proposal to import a biological control agent to Christmas Island will be regulated by the processes outlined in the Biosecurity Australia (Department of Agriculture, Fisheries and Forestry) document entitled Biosecurity Guidelines for the introduction of exotic biological control agents for the control of weeds and plant pests (http://www.daff.gov.au/ba/reviews/biological_control_agents/protocol_for_biological_control_ag ents/guidelines-introduction-exotic-bcas-weed-and-plants )

Issue 2 - Nomination of the lac scale Tachardina aurantiaca as a target species for a biological control agent

Step 1 of these guidelines is to have the target organism nominated as a candidate for biological control. In the referral we indicated that a nomination had been prepared but that the submission process was unclear because of jurisdictional issues surrounding Christmas Island an external territory of Australia (section 2.4, subsection entitled Regulatory Framework – Commonwealth or state legislation or policies under which approvals are required or will be considered against ). Resolution: That uncertainty has also been resolved. The Plant Health Committee, convened by DAFF, has responsibility for approving nominations of target species that are invertebrate pest species. The nomination of T. aurantiaca as a target species for a biological control agent was submitted to the committee and approved on 9 April 2013. Host Specificity Testing of Tachardiaephagus somervillei (Hymenoptera: Encyrtidae), a biological control agent for the yellow lac scale Tachardina aurantiaca (Hemiptera: Kerriidae)

G. Neumann, D.J. O'Dowd and P.T. Green

The yellow lac scale, Tachardina aurantiaca (Hemiptera: Kerriidae) is a damaging, invasive pest on Christmas Island and implicated in the formation of widespread, high-density supercolonies of the invasive yellow crazy ant ( Anoplolepis gracilipes ). Suppression and control of crazy ant supercolonies may be afforded by the importation and release of Tachardiaephagus somervillei (Hymenoptera: Encyrtidae), a widespread and abundant parasitoid of Tachardina that is known to parasitize lac scales in its native range in Southeast Asia (Hayat 2010, O’Dowd et al. 2012). Below we provide a preliminary evaluation of risk of importation and release of T. somervillei on Christmas Island and then describe a protocol for host-specificity testing.

Non-target issues on Christmas Island

To focus host-specificity testing for the exotic biological control agent Tachardiaephagus somervillei , we used three approaches to assess the risk its importation to and release on Christmas Island. First, we determined insect species on the island that could conceivably constitute non-target species. Records for occurrence of scale insect species (Superfamily Coccoidea, the same superfamily to which T. aurantiaca belongs) on the island were compiled from the literature and then supplemented by conducting >400 hours of structured search for endemic scale insects (O’Dowd et al. 2012). We then obtained the list of known endemic insect species on Christmas Island (James and Milly 2006) and narrowed consideration of taxa on this list to endemic species occurring in Order Hemiptera, the same order to which the known target species Tachardina aurantiaca belongs.

Second, we evaluated known host species records for all species of Tachardiaephagus using The Universal Chalcidoid Database (Noyes 2012) so that host range could be estimated and risk of release of T. somervillei assessed for species of concern on Christmas Island. Seven described species of Tachardiaephagus are distributed in Southeast and South Asia, and sub-Saharan Africa.

Third, we evaluated risk to endemic hemipterans by searching The Universal Chalcidoid Database for published records of the host ranges of encyrtid parasitoids known to attack members of the hemipteran families with endemic species on Christmas Island. The Encyrtidae (Hymenoptera, Superfamily: Chalcidoidea), the family to which Tachardiaephagus somervillei belongs, is one of the most important parasitic wasp (parasitoid) families for the biological control of harmful insects, including a variety of scale insects infesting woody plants (Noyes and Hayat 1994, Noyes 2012). The Encyrtidae currently comprises 460 genera and 3735 species in 2 subfamilies. The subfamily Encyrtinae includes 353 genera and 2920 species, while the Tetracneminae includes 107 genera and 815 species. Approximately half of all encyrtid species are associated with scale insects (Hemiptera: Coccoidea)(Noyes 2012). Encyrtids are generally endoparasitoids meaning that the parasitoid egg is laid directly inside the host’s body where the hatching larva completes development feeding on the host’s tissue, ultimately killing the host. Encyrtids mostly parasitize immature life stages (or, rarely, adults), but some species in one genus ( Microterys ) are egg predators (Noyes 2012). 2

The Universal Chalcidoidea Database (Noyes 2012) is the most comprehensive database for chalcidoid parasitoids, with over 120,000 host/associate records (including associations with food plants of the hosts) and > 140,000 distribution records of the parasitoids in the superfamily Chalcidoidea. It is very well developed, regularly updated and extremely well referenced. Nevertheless, large databases can contain errors affecting reliability, such as erroneous published host records and outdated parasitoid taxonomy (Kuhlmann et al. 2006). The most important source of error in this database is actual published erroneous records. Since all records are referenced, one can investigate a record that is in doubt. The database allows for searches for all recorded encyrtid associates of the hemipteran families represented by endemic species on Christmas Island. As a next step, all recorded primary hosts of encyrtid species associated with these taxa were determined. Consequently, any encyrtid species that share both coccoid species (scale insects) and any family represented by endemic species on Christmas Island can be found.

Evaluation of risk based host records of Tachardiaephagus and encyrtid parasitoids

Three lines of evidence indicate that there is a very low likelihood that importation and release of T. somervillei would harm species of concern on Christmas Island. First, all scale insects on Christmas Island that could be considered as the most likely potential non-target organisms are exotics and none are beneficial (Table 1). Nevertheless, the closest endemic relatives of the yellow lac scale on Christmas Island should still be considered to evaluate the risk of becoming non-target hosts of T. somervillei . These endemic species comprise seven species in the Order Hemiptera - a true bug, a cicada, a leafhopper, a spittlebug, and three planthoppers (Table 2). All of these endemic species occur in different suborders (either suborder Auchenorrhyncha or Heteroptera) than the yellow lac scale (suborder Sternorrhyncha). Thus, any potential for non- target impacts by T. somervillei would require a host range that bridges this very substantial phylogenetic distance (see Fig. 2, Cryan and Urban 2012), life-history differences, and ecological distinctiveness.

Second, host records of Tachardiaephagus strongly suggest that all species are specialists on lac scales in the Kerriidae (Table 3). All host species records of Tachardiaephagus (Kerria , Tachardina , and Paratachardina ) are restricted to the lac scale family Kerriidae. Furthermore, all host species records for T. somervillei are within two genera: Kerria , the lac insect of commerce, and Tachardina . Thus, it is highly likely that T. somervillei has a very narrow host range restricted to kerriid host species. Tachardiaephagus tachardiae is reported to use Paratachardina lobata (Kerriidae) in India (Table 3). The only non-target species on Christmas Island that T. somervillei would have a likelihood to use is the false lobate lac scale Paratachardina pseudolobata (Kerriidae)(Table 1), which is native to Southeast Asia and introduced and invasive on Christmas Island. P. pseudolobata is a damaging invader elsewhere in its introduced range (Schroer et al. 2008).

Third, the broader analysis of host records for primary parasitoids in the Encyrtidae, the family to which T. somervillei belongs, shows that there are no known records in the Encyrtidae of any parasitoid species that attacks both scale insects (Coccoidea) and host species in hemipteran families that have endemic representatives on Christmas Island (Table 2). All known encyrtid primary parasitoids known to attack species in families which have endemic representatives on 3

Christmas Island have primary hosts only within suborder Auchenorrhyncha. While species in these hemipteran families are attacked by many species of chalcidoid parasitoids, many times fewer or no encyrtid species are known to use them as hosts, and not one encyrtid species is known with a host range so extreme that it encompasses both scale insects and any of these hemipteran families.

Host specificity testing protocol

Actual records for host range of Tachardiaephagus and known patterns of parasitism in the Encyrtidae indicate that it is highly improbable that T. somervillei would exhibit an extremely broad host range unprecedented in Encyrtidae and utilize any of the endemic hemipteran species on Christmas Island. Thus, when the host specificity testing of the biological control agent is initiated using the centrifugal phylogenetic approach, the outer boundaries of the sequential testing should be set initially at a significantly closer distance from the target organism than the taxa represented on Christmas Island by endemics.

Terminology . A ‘ test species ’ is an insect species that is tested as a non-target insect. Since no non-target scale insects were identified on Christmas Island, the list of test species will be determined using the centrifugal approach (Kuhlmann et al. 2006, Neumann et al. 2010). In this approach, species other than the known target host of the biological control agent are tested with the most closely related species (least phylogenetic distance, and with close similarities in biology and ecology) tested first then less similar species thereafter (“centrifugal principle”).

A test species will be considered a ‘ suitable host ’ if parasitoid emergence is observed during any of the exposure tests. To be a suitable host, the parasitoid must (a) accept the test species when exposed, (b) must oviposit, (c) the eggs must hatch into larvae, (d) the larvae must complete development using the test species as the resource, (e) the larvae must pupate and, finally, (f) the emerging adult parasitoid must be able to exit (emerge from) the test species. If parasitoid emergence from a test species is observed, the test species will be considered a suitable host without assessing the viability, fecundity, sex ratio and other characteristics of the emerging parasitoid generation.

A test species will be considered ‘ susceptible’ if the parasitoid causes significant mortality due to probing, host feeding (when the parasitoid stabs the insect with its ovipositor and then feeds on the insect’s hemolymph), oviposition, or oviposition and larval development. Note that even if some mortality is observed due to the above, the test species will only be considered susceptible if the mortality resulting to the exposure to the parasitoid is significantly higher than in control groups (negative controls) of the test species where the test individuals are not exposed to the parasitoid.

The ‘ parasitoid’ in this case is Tachardiaephagus somervillei (Hymenoptera: Encyrtidae), a widespread parasitoid native to Southeast Asia. It is known to parasitize some scale insects in the family Kerriidae (lac scales) including the ‘ target host ’, Tachardina aurantiaca (O’Dowd et al. 2012) .

Selecting test species. Host specificity testing can be a long and difficult process, depending on how problematic it is to establish and manage colonies of the parasitoid, the target host, and the 4 test species. Furthermore, resources and time are not unlimited and permitting issues to collect insects and run field experiments (when needed) can add to the time required to complete host testing. Therefore, a manageable number of species should be used (10 to 15, but fewer are acceptable if high specificity is found initially). The species used will be largely determined by availability and determining the test species at species or genus level is not possible at this point. We can, however, predict to some degree the families of the test species. We will focus on neococcid taxa including the Kerriidae, the family to which the target host belongs (see Gullan and Cook 2007). Considering the phylogenetic relationships of scale insect families (Fig. 1), we will aim to test more than one species from the family Coccidae. Species in the family Diaspididae will also be considered as a less closely related group of scale insects. Early in the host testing process, an ‘out-group’ (test species phylogenetically more distant) will also be used, most likely selected from the family Pseudococcidae (Fig. 1).

As mentioned above, availability will determine species selection to some degree but in some of the groups (e.g. Coccidae) there may be a number of species available. Species that can be reared in laboratory conditions would be preferred but carrying out tests in field conditions is also acceptable as long as the tests can be well controlled.

No-choice tests. In no-choice tests, only the non-target test species will be provided to the parasitoid in the experimental replicates. This is the most stringent test method that places the parasitoid under high “oviposition pressure”.

For each test species, trials will be replicated 10 times. Each replicate will consist of 1) one experimental cage with the test species (50 individuals in three different age groups/instars; 150 individuals in total) exposed to 10 female and 10 male parasitoids, 2) one positive control cage consisting of 50 mature female T. aurantiaca exposed to 5 female and 5 male parasitoids in order to confirm that the parasitoids used in the experiment are of good quality and capable of parasitism in the given exposure time, and 3) one negative control cage which will be similar to the experimental cage except the test species will not be exposed to parasitoids. The negative controls will be used to determine any effect of parasitoid exposure on insect mortality other than successful parasitization such as probing, host feeding and oviposition without successful parasitoid development.

The cages will consist of fine mesh bags (sleeves) fixed on the branches of plants that serve as host plants for the scale insects. Parasitoids will be collected from a lab colony and introduced to the cages < 24 h after emergence. Exposure time to parasitoids will be determined during preliminary studies to determine the optimal time frame for mating, egg maturation and oviposition by the parasitoid.

Evaluation:

1) If the positive control does not yield any parasitoid emergence, the replicate will not be evaluated (failed replicate). 2) If the positive control yields parasitoid emergence the replicate will be evaluated. The parasitization rate in the positive will be noted but will not be compared to parasitization rates in the test replicate. 5

3) Parasitoid emergence in the test replicate will be noted along with the approximate age and developmental stage of the individuals yielding parasitoid emergence. These individuals will not be used to compare mortality rates due to reasons other than successful parasitism. 4) Mortality of test species in test replicates and negative controls will be noted.

Analysis:

1) If any test replicates yielded any number of parasitoid emergences, the test species will be considered a suitable host. The level of suitability of the test species will not be compared to the suitability of the target host at this stage of the project. 2) Mortality (except mortality due to successful parasitism) rates of test insects and insects in the negative control will be compared using 2-sample t-test. If the mortality rate is significantly higher (α = 0.05), the test species will be considered susceptible and the reason for mortality will be investigated.

If a test species is susceptible (but not suitable), behavioural observations will be made to determine the cause of the effect of parasitoid exposure. The observational arena will be a ‘window box’ arena. This arena is constructed by constructing a 10-cm W x 10-cm L x 5-cm H box with one side being glass. The box is constructed so that it can be placed on the stem of a potted host plant on which the test insects feed. Female parasitoids are then liberated inside the box and the box closed. The potted plant then can be placed flat on its side and the box positioned under a dissecting microscope. Video footage is then captured using a high definition video camera with a microscope adaptor. The footage (4-6 h per observation) can be analyzed later for the cause(s) of mortality (i.e. probing, host feeding, or oviposition without full parasitoid development). If oviposition is suspected, dissections will also be made to investigate parasitoid larval development.

Other testing methods . We considered using choice tests (where the parasitoid is presented with the test species and the target host simultaneously) and sequential no-choice tests (where the parasitoid is presented with the target host first then it is transferred on to the test species). However, in our case these tests may have limited value.

A choice test would show whether a test species that is a suitable host in no-choice tests would be consistently attacked when the target host is also available. A physiologically suitable host may not be preferred over the target host and parasitization would not occur (or at very low level) in field conditions. This test would be used in cases where the suitable host (found suitable in a no-choice test) is of concern, e.g. a beneficial or endemic non-target species that coexists with the target host. However, this situation does not arise on Christmas Island where all known scale insects are exotic and none of them are beneficial. This test is further limited by the assumption that the test species and the target host occupy the same space hence providing the parasitoid with a choice. This would also not be the case on Christmas Island.

A sequential no-choice test would further show the unsuitability of a test species. The parasitoid may be ‘motivated’ to attack the test species even if it did not attack it in a simple no-choice test if the parasitoid is first allowed to initiate oviposition behavior and to gain experience. This test may be even more stringent than the simple no-choice test and one would consider using it in the 6 case of a test species that is of great concern to further show the unsuitability of the test species. This would again require a beneficial or endemic non-target species. Furthermore, similar to the choice test, this would assume either the coexistence or very close proximity of the test species and the target host.

Options for the location for host-specificity testing

There are three options for the location where host specificity testing could be conducted: 1) in quarantine containment at the release location (Christmas Island); 2) in quarantine containment on mainland Australia; and 3) in the native geographical range of the biological control agent where no containment would be necessary.

Option 1. Testing in containment at the release location (Christmas Island)

This option is the least attractive and involves increased risks and expenses. There is no quarantine containment facility on Christmas Island, a remote oceanic island. The expense of constructing such a facility to Quarantine Approved Premises Criteria (Quarantine Insectary Level 2) for host-specificity testing of a single agent would be prohibitive. Even if there were such a facility, in case of escape, the biological control agent would find an environment very similar to its native range with its natural host T. aurantiaca in abundance. The benefits of this option would include easy access to the natural host for the parasitoid colony and no further travel if and when a release permit is obtained.

Option 2. Testing in containment on mainland Australia

While this option is more attractive than Option 1, it is still somewhat risky and labour and cost intensive. It brings few benefits. The biological control agent (parasitoid) would have to be brought into containment along with its natural host ( Tachardina aurantiaca ). This would pose the risk of not only a parasitoid escaping but also a potentially invasive, host-generalist scale insect (on Christmas Island, for example, at least 15 horticultural species are attacked by T. aurantiaca , including three species of Citrus, Macadamia, Guava, Pomegranate, Chili, Eggplant, Star fruit, and Soursop) (R.W. Pemberton and D.J. O’Dowd, unpublished results). Using a containment facility in temperate region could decrease the risks. In case of quality control issues of the biological control agent due to ‘lab selection,’ additional travel, collection, and importation of the biological control agent would be necessary, further increasing costs. Natural host colony loss (which can easily happen due to overexposure to the parasitoid agent, fungal infections, etc.) would necessitate multiple importations of the natural host, which would once again significantly increase costs and risk.

Option 3. Testing in the native geographical range of the biological control agent with no containment necessary

In our case, this option is the most attractive, most cost-effective, and least risky. There are many benefits to study biological control agents and carry out host specificity studies in the native geographical range of the agent. In our case, both the parasitoid and its natural host is readily available from the wild in Malaysia which ensures good quality parasitoid and host cultures. The risk factor in our case would be zero as there would be no importation of any organisms. Although obviously desirable, such studies in the native range are not common, in part because 7 of the lack of local research facilities, lack of skilled and reliable cooperators, or sufficient time in what are sometimes difficult locations (Van Driesche et al. 2008). Scientific work is cost-effective in Malaysia compared to Option 2 and our network of collaborators is well established (see Table B1, O’Dowd et al. 2012). Options 1 and 2 offer no benefits over this option. We therefore suggest that host specificity testing is carried out in the native range of the biological control agent. This option carries the least risk and is consistent with Recommendation 2 in a DAFF- commissioned review of biosecurity risks in biological control (Ferrar et al. 2004) that biological control practitioners undertake host specificity testing in the native region before any importation.

References cited

Ben-Dov, Y., Miller, D.R. & Gibson, G.A.P. 2012. Scalenet ( http://www.sel.barc.usda. gov/ scalenet/scalenet.htm).

Cryan, R.C. and J.M. Urban. 2012. Higher-level phylogeny of the insect order Hemiptera: is Auchenorrhyncha really paraphyletic? Systematic Entomology 37: 7-21.

Ferrar, P., I.W. Forno and A.L. Yen. 2004. Report of the Review of the Management of Biosecurity Risks Associated with the Importation and Release of Biological Control Agents. Australian Government Department of Agriculture, Fisheries and Forestry, Canberra, Australia. 27 pp.

Gullan, P. J. and L. G. Cook. 2007. Phylogeny and higher classification of the scale insects (Hemiptera: Sternorrhyncha: Coccoidea). Zootaxa 1668: 413-425.

Hayat, M., S. Schroer, and R.W. Pemberton. 2010. On some Encyrtidae (Hymenoptera: Chalcidoidea) on lac insects (Hemiptera: Kerriidae) from Indonesia, Malaysia and Thailand. Oriental Insects 44: 23-33.

James, D. and N. Milly. 2006. A biodiversity inventory database for Christmas Island National Park. A report for the Department of Finance & Administration and Department of Environment & Heritage. Director of National Parks, Australian Government, Canberra. 49 pages.

Kuhlmann, U., U. Schafner and P. G. Mason. 2006. Selection of non-target species for host specificity testing. In: Environmental Impact of Invertebrates for Biological Control of Arthropods: Methods and Risk Assessment (eds. F. Bigler et al.). CAB International.

Neumann, G., P. A. Follett, R. G. Hollingsworth, and J. de León. 2010. High host specificity in Encarsia diaspidicola (Hymenoptera: Aphelinidae), a biological control candidate against the white peach scale in Hawaii. Biological Control 54: 107-113.

Noyes, J.S. and M. Hayat. 1994. Oriental mealybug parasitoids of the Anagyrini (Hymenoptera: Encyrtidae) viii+554pp. CAB International, Wallingford, UK.

Noyes, J.S. 2012. Universal Chalcidoidea Database. World Wide Web electronic publication. http://www.nhm.ac.uk/chalcidoids 8

O’Dowd, D.J., P.T. Green, G. Neumann, and S. Wittman. 2012. Executive summary: research and development for indirect biological control of the yellow crazy ant ( Anoplolepis gracilipes ) on Christmas Island, Indian Ocean. Unpublished report to the Director of National Parks and the Crazy Ant Scientific Advisory Panel, 27 pp.

Prinsloo, G. 1983. Entomological Memoirs of the Department of Agriculture Republic of South Africa 60: 26.

Schroer, S., Pemberton, R.W., Cook, L.G., Kondo, T., and Gullan, P.J. 2008. The genetic diversity, relationships, and potential for biological control of the lobate lac scale, Paratachardina pseudolobata Kondo & Gullan (Hemiptera: Coccoidea: Kerriidae). Biological Control 46: 256-266.

Van Driesche, R., T. Center, M. Hoddle and N. Mills. 2008. Can efficacy of new biological control agents be predicted before their release? Pp. 1-13 in Mason, P.G., D.R. Gillespie and C. Vincent (eds). Proceedings of the Third International Symposium on Biological Control of Arthropods. Christchurch, New Zealand, 8-13 February 2008. United States Department of Agriculture, Forest Service, Morgantown, WV, FHTET-2008-06, December 2008, 636 p. 9

Table 1. Scale insects of Christmas Island. It is highly probable that all of these species, with broad host plant ranges and geographic distributions, are exotic to Christmas Island and introduced following human settlement. Taxonomy and distribution from Ben-Dov et al. (2012, http://www.sel.barc.usda. gov/ scalenet/scalenet.htm).

Family and Species Common Name Distribution

Kerriidae (lac scales) Tachardina aurantiaca (Cockerell) Yellow lac scale Oriental Paratachardina pseudolobata False lobate lac scale Oriental, Nearctic, Neotropical (Kondo & Gullan) Coccidae (soft scales) Coccus celatus De Lotto Green coffee scale Afrotropical, Australasia, Oriental *C. hesperidium Linnaeus Brown soft scale Cosmopolitan *Milviscutulus mangiferae (Green) Mango shield scale Cosmopolitan *Ceroplastes ceriferus (Fabricius) Indian wax scale Cosmopolitan *C. destructor Newstead White wax scale Afrotropical, Australasia, Oriental Saissetia oleae (Olivier) Black olive scale Pantropical *S. coffeae (Walker) Hemispherical scale Cosmopolitan ** Parasaissetia nigra (Nietner) Nigra scale Cosmopolitan ** Pulvinaria urbicola Cockerell Urbicola soft scale Pantropical ** P. psidii 1 Green shield scale Cosmopolitan Monophlebidae (giant scales) Icerya purchasi (Maskell) Cottony cushion scale Cosmopolitan

Cerococcidae (ornate pit scales) ** Cerococcus indicus (Maskell) Spiny brown coccid Cosmopolitan

Pseudococcidae (mealybugs) ** Dysmicoccus finitimus Williams Asian coconut Australasia, Oriental mealybug ** Ferrisia virgata (Cockerell) Striped mealybug Cosmopolitan

Diaspididae (armoured scales) Aspidiotus destructor (Signoret) Coconut scale Cosmopolitan Pseudaulacaspis pentagona White peach scale Cosmopolitan (Targioni Tozzetti) *Hemiberlesia palmae (Cockerell) Tropical palm scale Cosmopolitan *Lindingaspis sp. -- -- ** Ischnaspis longirostris 1 Black thread scale Cosmopolitan

*Record added by Abbott (2004, PhD Thesis, Monash University); ** Records added by Neumann et al. (unpublished results); 1Tentative identifications 10

Table 2. Endemic hemipteran species known from Christmas Island (James and Milly 2006) and primary parasitoids (superfamily: Chalcidoidea, family: Encyrtidae) associated with the families represented by the endemic species. The data were extracted from the Universal Chalcidoidea Database (Noyes 2012). The families of Nogodonidae and Rhopalidae have no associated chalcidoid primary parasitoids and therefore endemic species in these families on Christmas Island can most likely be excluded from all further consideration. Cicadellidae and Delphacidae have the highest diversity of chalcidoid primary parasitoids but have magnitudes lower diversity of encyrtid primary parasitoids. These data indicate that none of the encyrtid primary parasitoids of families with endemic species on Christmas Island have an overlapping host range with the higher taxa (suborder Sternorrhyncha) that includes the target lac scale Tachardina . During the database analysis, only records with species-level chalcidoid identification were used. N/A indicates not applicable.

No. chalcidoid No. encyrtid 1o Suborder/Family host associates parasitoid range of encyrtids Endemic species Family of family species of family parasitizing family

Xestocephalus Cicadellidae 627 6 Auchenorrhyncha izzardi (Cicadellidae) Oxypleura Cicadidae 35 0 N/A calypso Clovia eximia Cercopidae 71 4 Auchenorrhyncha (Cercopidae- Aphrophoridae) Ugyops aristella Delphacidae 248 5 Auchenorrhyncha (Delphacidae- Cicadellidae) Varcia Nogodinidae 0 N/A N/A flavicostalis Salona oceanica Leptocoris Rhopalidae 0 N/A N/A subrufescens

11

Table 3. Records of known host families and genera for the primary parasitoid Tachardiaephagus (Encyrtidae). The biological control agent under investigation, Tachardiaephagus somervillei , is in bold. As a genus, Tachardiaephagus has an extremely broad geographic range. With the exception of one probably erroneous host record in Africa, all Tachardiaephagus species appear to be family specialists and restricted to the Kerriidae. For host genera, number of species recorded as hosts is in parentheses. Based on Noyes (2012, Universal Chalcidoidea Database, http://www.nhm.ac.uk/ research-curation/research/projects/chalcidoids/database/), except for records for additional records for T. somervillei and T. sarawakensis (Hayat et al. 2010; O’Dowd et al. 2012; R.W. Pemberton, pers. comm.)

Species Distribution Recorded hosts

Tachardiaephagus India, Malaysia, Thailand Kerriidae somervillei Kerria spp. (4) Tachardina sp. 1 T. sarawakensis Sarawak Kerriidae Tachardina aurantiaca T. tachardiae Brunei, China, India, Indonesia, Kerriidae Malaysia, Sri Lanka, Taiwan, Kerria spp. (8) Vietnam, Azerbaijan Paratachardina lobata T. similis Afrotropical, South Africa Kerriidae Tachardina sp. (1) T. absonus Afrotropical, South Africa Kerriidae Tachardina spp. (2) T. communis Afrotropical, South Africa Kerriidae Tachardina spp. (5) T. gracilis Afrotropical, South Africa Kerriidae Tachardina sp. (1)

1 Probably T. aurantiaca , since it is the only known Tachardina species in Asia.

12

Pseudococcidae

Coccidae

Kerriidae Kermesidae Diaspididae

Gondwanan clade

BSE clade

Asterolecaniidae Acanthococcid group

Dactylopiidae

Figure 1. Portion of a phylogram of the scale insects (Coccoidea) based on nucleotide sequencing of nuclear 18S rRNA for 72 species of scale insects and 10 outgroup taxa (see Gullan and Cook [2007] for details). Only the neococcoids (red) are shown here. For all families given here (except the Kermesidae), Bayesian posterior probabilities (numerical values above branches) are >90%. The lac scales Kerriidae (bold italics) is a sister family of the soft scales Coccidae, more distant to the armoured scales Diaspididae, and even further removed from the mealybugs Pseudococcidae.

Nomination of the Yellow lac scale as a biological control agent target

Target: Yellow lac scale Tachardina aurantiaca Cockerell

Nominating Organisation:

Parks Australia, Department of Sustainability, Environment, Water, Populations and Communities

Prepared by Gabor Neumann, Peter Green, and Dennis O’Dowd La Trobe University Bundoora, Vic 3086

1 1. Taxonomy Class: Insecta Order: Hemiptera Sub-order: Sternorrhyncha Superfamily: Coccoidea Family: Kerriidae Subfamily: Tachardininae Genus: Tachardina Species: aurantiaca Cockerell

Common names: Yellow lac scale, Golden lac scale, Mooncake lac scale

Synonyms: Not applicable

2. Description Female scales usually separate, sometimes coalescing, round when seen from above, 4 mm. long, convex, but flattened dorsally, so that they are not half as high as broad; surface thrown more or less into concrete folds; colour bright yellow; median dorsal area ferruginous, with radiating ridges and the usual orifices, the minutely transversely ribbed larval exuviae in the middle. Young up to about 2 mm long, orange-ferruginous, with rather obscure radiating ridges. Resinous test of adult female (Fig. 1) circular, somewhat flattened dorsally: the larval pellicle forming a crenulated ridge in the centre of the dorsal area: anal orifice circular or broadly oval, its posterior rim raised into a prominent tooth- like point: respiratory orifices small, very slightly prominent, situated one on each side of and close to the larval pellicle: sides more or less distinctly broadly radially fluted. Colour bright fulvous to castaneous, the larval pellicle reddish; semitranslucent (Ben-Dov 2012). The test (cover) of males is elongate and red or reddish brown with an operculum at the posterior end through which the winged, red males emerge. Males are short-lived and may not play an important role in reproduction (Fig. 2).

Fig. 1. Mature females of the yellow lac Fig. 2. Male tests of the yellow lac scale scale tended by the yellow crazy ant and an emerged male. Photo: Ong Su Ping Anoplolepis gracilipes . Photo: S. Belcher

2 3. Distribution

3.1 Native geographic range and centre of origin Southeast Asia: Indonesia (Java); Malaysia (Peninsular Malaysia, Sarawak, Sabah); Singapore; Thailand (Ben-Dov 2012). Likely centre of origin is Sundaland in Southeast Asia.

3.2 Australian and overseas distribution Indian Ocean islands: Christmas Island, Indian Ocean (External territory of Australia) (Campbell 1964; Ben-Dov 2012); Maldives, Indian Ocean (Ben-Dov 2012)

4. Proposing Organisation Parks Australia, Christmas Island, WA 6798 La Trobe University, Bundoora, VIC 3086

5. Pest status and costs The yellow lac scale is an invasive environmental pest on Christmas Island (Indian Ocean). A broad host plant generalist, it attacks at least 29 native plant species on the island (Abbott 2004). Its mutualistic association with the exotic yellow crazy ant Anoplolepis gracilipes leads to population build-up of both species and large impacts on the rainforest ecosystem - so-called invasional meltdown (O’Dowd et al. 1999; O’Dowd et al. 2003). Direct effects include the rapid decline of the dominant rainforest tree Inocarpus fagifer and widespread forest dieback. Indirect impacts include the population collapse of the dominant keystone species (the red land crab) on the island (O’Dowd et al. 2003), negative effects on endemic land birds (Davis et al. 2008, 2010), and invasion and rapid spread of the giant African land snail Achatina fulica in island rainforest (Green et al. 2011). Invasion by yellow crazy ants on Christmas Island has been Listed since 2005 as a Key Threatening Process under the EPBC Act 1999 (DSEWPaC 2005). Furthermore, the association between honeydew-secreting scale insects and invasive ants was recognized as a key threat to biodiversity on Christmas Island (Christmas Island Expert Working Group 2010), and the Australian government accepted the recommendation of the Expert Working Group to accelerate biological control efforts for scale insects on Christmas Island (Commonwealth of Australia 2011).

Current annual recurrent costs of mitigating the damage caused by the association of scale insects with the yellow crazy ant is one million dollars per year (i.e. allocated funding of $4 million over four years from 2011-12 to 2014- 15).

6. Beneficial status The yellow lac scale is a major environmental pest in its area of introduction and invasion on the Australian External Territory of Christmas Island, and is not known as a beneficial insect in its centre of origin.

3 7. Control methods

7.1 Chemical control Although no direct chemical control methods have been developed for the yellow lac scale, topical and systemic insecticides have been trialled for a related lac scale Paratachardina pseudolobata that is invasive in southern Florida, USA (Howard and Steinberg 2005). These methods require either multiple foliar applications of insecticide (bifenthrin) or root drenching of individual host plants (imidacloprid). Neither treatment methodology is logistically feasible for T. aurantiaca , which occurs in the canopies of 30-metre tall trees that are distributed across 100 km 2 of tropical rainforest in a national park on a remote and rugged oceanic island. Even if it were feasible, these insecticides are non- specific insect neurotoxins that could generate non-target impacts on native insects (e.g. Krischik et al. 2007). Population outbreaks of T. aurantiaca are suppressed indirectly when its tending ant mutualist, the yellow crazy ant, is controlled with toxic bait (Abbott and Green 2007, Green and O’Dowd 2010). However, yellow lac scale and crazy ant populations frequently recover in baited areas and form new infestations elsewhere, necessitating repeated chemical control for more than a decade. Moreover, the toxic bait containing fipronil has some undesirable non-target impacts.

7.2 Biological control On Christmas Island, the yellow lac scale is attacked by one hymenopteran associate and two species of scale predatory Lepidoptera (Table 1). Marietta leopardina is a parasitoid that exclusively attacks male yellow lac scales, but given the continued high densities of lac scales in supercolonies of invasive crazy ants, it is clearly not an effective biological control agent. The two lepidopteran predators (Eublemma sp. and ? Holcocera sp.) are rare and do not seem to have any significant effect on yellow lac scale populations on the island.

Table 1. Insect biological control agents of the yellow lac scale Tachardina aurantiaca on the Australian external territory of Christmas Island.

Order and family Species

Hymenoptera: Encyrtidae Marietta leopardina Motschulsky

Lepidoptera: Noctuidae Eublemma sp. (tentative) Blastobasidae Holcocera sp. (tentative)

8. Potential Conflicts of Interest There are no known conflicts of interest associated with the proposed biological control of the yellow lac scale. Its only known impacts on Christmas Island are negative, and it is not an important food item for any native species. Its association with the exotic yellow crazy ant constitutes a threat to the biodiversity and conservation values of the Christmas Island National Park (Christmas Island Expert Working Group 2010, Commonwealth of Australia 2011). Furthermore, the cost of mitigating the damage caused by the association between the yellow lac scale and the yellow crazy ant has cost the Australian government millions of dollars since baiting commenced in 1999.

4

References

Abbott, K. (2004) Alien ant invasion on Christmas Island, Indian Ocean: the role of ant- scale associations in the dynamics of supercolonies of the yellow crazy ant Anoplolepis gracilipes . PhD thesis, Monash University, Melbourne, Victoria, Australia.

Abbott, K. and Green, P.T. (2007). Collapse of ant-scale mutualism in a rainforest on Christmas Island. Oikos 116, 1238-1246.

Ben-Dov, Y. (2012). ScaleNet, Tachardina aurantiaca . 13 June 2012. http://www.sel.barc.usda.gov/catalogs/kerriida/Tachardinaaurantiaca.htm

Campbell, T.G. (1964). Entomological survey of Christmas Island (Indian Ocean) with special reference to the insects of medical, veterinary, agricultural and forestry significance. CSIRO, Division of Entomology, Canberra, Australia. 48 pp.

Christmas Island Expert Working Group (2010). Final report of the Christmas Island Expert Working Group to the Minister for Environment, Heritage and the Arts. http://www.environment.gov.au/parks/publications/christmas/pubs/final-report.pdf

Commonwealth of Australia (2011). Australian government response to the recommendations of the Christmas Island Expert Working Group. http://www.environment.gov.au/parks/publications/christmas/pubs/ewg-report- response.pdf

Davis, N.E., O’Dowd, D.J., Green, P.T., and Mac Nally R.T. (2008) Effects of an alien ant invasion on abundance, behavior, and reproductive success of endemic island birds. Conservation Biology 22, 1165-1176.

Davis, N.E., O’Dowd, D.J., Mac Nally R.T., and Green, P.T. (2010) Invasive ants disrupt frugivory by endemic island birds. Biology Letters 6, 85-88.

DSEWPaC (2005). Loss of biodiversity and ecosystem integrity following invasion by the Yellow Crazy Ant ( Anoplolepis gracilipes ) on Christmas Island, Indian Ocean. http://www.environment.gov.au/cgi-bin/sprat/public/publicshowkeythreat.pl?id=16

Green, P.T. and O’Dowd, D.J. (2010). Management of invasive invertebrates: lessons from the management of an invasive alien ant. Pp. 153-172 in (Clout, M. and P. Williams, eds.) Invasive species management: A Handbook of Principles and Techniques . Oxford University Press, Oxford.

Green, P.T., O’Dowd, D.J., Abbott, K., Jeffrey, M., Retallick, K., and Mac Nally, R.T. (2011). Invasional meltdown: Invader–invader mutualism facilitates a secondary invasion. Ecology 92 , 1758-1768.

Howard, F.W. and Steinberg, B. (2005). Root drenches and topical insecticide treatments for control of the lobate lac scale, Paratachardina lobata (Chamberlin). Florida State Horticultural Society 118, 314-318.

Krischik, V., Landmark, A., and Heimpel, G. (2007). Soil-applied imidacloprid is translocated to nectar and kills a nectar-feeding Anagyrus pseudococci (Girault) (Hymenoptera: Encyrtidae). Environmental Entomology 36,1238-1245.

5 O’Dowd, D.J., Green, P.T., and Lake, P.S. (1999) Status, impact, and recommendations for research and management of exotic invasive ants in Christmas Island National Park. Report to Environment Australia. http://www.issg.org/database/species/reference_files/Christmas_Island_Report.pdf

O’Dowd, D.J., Green, P.T., and Lake, P.S. (2003). Invasional ‘meltdown’ on an oceanic island. Ecology Letters 6, 812-817.

6 Executive summary

Research and development for indirect biological control of the yellow crazy ant (Anoplolepis gracilipes) on Christmas Island, Indian Ocean

Dennis J. O’Dowd, Peter T. Green, Gabor Neumann, and Sarah Wittman Department of Botany, La Trobe University, Bundoora, Vic 3086

Issue: Supercolonies of the yellow crazy ant (YCA) are a major and on-going threat to biodiversity values on Christmas Island. To date their management has depended on surveillance, monitoring, and control using toxic bait. Nevertheless, new supercolonies continue to form. There is widespread concern for the sustainability of this program in terms of its expense, non-target impacts, and the resources it diverts from other conservation programs. Long-term, sustainable suppression of YCA supercolonies could be achieved through the introduction of a biological control agent that would indirectly affect YCA by reducing carbohydrate supply provided by scale insects, a key resource implicated in supercolony dynamics.

Project goal: Establish the feasibility of suppressing YCA supercolonies and their impacts through safe biological control of its scale insect mutualist, the yellow lac scale Tachardina aurantiaca (Kerriidae)

Project objectives:

1. Evaluate the concept of indirect biological control by assessing the dependence of YCA supercolonies on carbohydrate supply (Section A below and supporting figures A1-A7). 2. Identify prospective biological control agents (BCA) of Tachardina aurantiaca and evaluate the risk of their introduction to Christmas Island (Section B below and supporting tables B1-B4 and figures B1-B4).

Background Information: Invasion by yellow crazy ants on Christmas Island has been listed since 2005 as a Key Threatening Process under the EPBC Act 1999 (DSEWPaC 2005). Moreover, the association between exotic honeydew-secreting scale insects and invasive ants has been recognized as a key threat to biodiversity on Christmas Island (Christmas Island Expert Working Group 2010). The Australian government accepted the recommendation of the Expert Working Group to accelerate biological control efforts for scale insects on Christmas Island (Commonwealth of Australia 2011). The research described below is funded by Parks Australia and conducted by La Trobe University under a contract by the Director of National Parks and with the advice of the Crazy Ant Scientific Advisory Panel, an advisory group of scientists and managers for the control of the yellow crazy ant on Christmas Island.

1

A. Role of Carbohydrates in YCA supercolony dynamics

A1. Carbohydrate supply and YCA laboratory colony dynamics and behaviour

Key outcome: Dynamics and behaviour of YCA in laboratory colonies depended on carbohydrate supply. When sugar supply was elevated, reproductive output by queens increased, death rates of workers decreased, foraging tempo quickened, and interspecific aggression intensified (Figs. A1 and A2). These results suggest that sugar supply, through honeydew production by scale insects, plays an important role in YCA supercolony dynamics, foraging efficiency, and interspecific aggression.

A2. YCA supercolony status and carbohydrate supply

Key outcome: Stable isotope analyses (δ15N) of YCA from four declining supercolonies showed that trophic position, a measure of diet composition, increased (i.e. YCA became more carnivorous) as population densities decreased over 18 months in 2010-2011. This result suggests that a waning carbohydrate supply, as indicated by an upward shift in the trophic position of YCA, is associated with supercolony decline (Fig. A3). A series of YCA collected over two years from 10 supercolonies in 2000-2002 were also analyzed for δ15N after storage in ethanol for over a decade. Unlike the four sites above, δ15N values and YCA abundance were not negatively correlated at three supercolonies that declined to zero over the period. However, as expected, δ15N values for the other seven supercolonies with consistently high YCA densities did not vary over time.

A3. Field exclusion of YCA from honeydew resources in the forest canopy.

Key outcome: Denying YCA access to honeydew resources in the forest canopy resulted in a rapid decline in their abundance (Figs. A4 and A5). Tree bands effectively denied YCA access to the canopy and led to a rapid decline in YCA abundance on the forest floor. Foraging activity on the forest floor fell four-fold within four weeks. This field experiment, conducted at an appropriate spatial scale, confirms the importance of carbohydrate supply to supercolony dynamics suggested by both the laboratory experiment (A1) and the stable isotope analyses (A2).

A4. Relative importance of Tachardina aurantiaca in the honeydew economy of YCA supercolonies.

Key outcome: Successful biological control of the yellow lac scale Tachardina aurantiaca could remove a large fraction of honeydew available to the yellow crazy ant (Fig. A6). Calculation of a site honeydew index based on forest stand structure and composition, and the capacity of different tree species to host important honeydew-producing scale species (T. aurantiaca or coccoid soft scale species) indicated that the yellow lac scale contributed an estimated average of 70% (range 46-86%) of the total honeydew economy on ten 0.25 ha forest stands that supported YCA supercolonies in 2000-2002). However, these estimates are likely to have decreased in some supercolonies over the past decade as a result of the decline of Inocarpus fagifer, a key host plant of T. aurantiaca (Fig. A7). These issues may constitute the largest risk to the success of the program.

2

B. Research and survey for prospective biological control agents of the yellow lac scale Tachardina aurantiaca

B1. Establishment of co-operators to facilitate research and development of biological control for Tachardina aurantiaca

Key outcome: A network of co-operators and collaborators is now established in Australia, India, Southeast Asia, and the United States to facilitate survey, identification, and culture of prospective agents for the biological control Tachardina aurantiaca on Christmas Island (Table B1).

B2. Scale insect survey for risk assessment of introduction of a biological control agent to Christmas Island

Key outcomes: No native or endemic scale insect species have been discovered in intensive and extensive searches for scale insect species on Christmas Island (Fig. B1). However, 400 hours of search over two years did yield five additional exotic scale insect species previously unknown to the island (Table B2). Assuming that any prospective biological control agent for Tachardina aurantiaca on Christmas Island would be a narrow family specialist (Kerriidae) and all known scale insect species are exotic/invasive, the probability of any direct non-target effects is negligible.

B3. Search for Tachardina aurantiaca in its native range, Southeast Asia

Key outcomes: Populations of the yellow lac scale Tachardina aurantiaca have been located in Malaysia, within its native range in Sundaland (Southeast Asia). Nine populations, all small and isolated, were located: four in Peninsular Malaysia and Singapore, and five in Malaysian Borneo (Sarawak and Sabah)(Fig. B2). Cuticular morphology and mitochondrial COI and S28 ribosomal RNA sequences are identical among T. aurantiaca on Christmas Island, and preliminary analyses indicate that Malaysian T. aurantiaca are morphologically and genetically identical to those on Christmas Island. Thus, the single population of T. aurantiaca on Christmas Island appears to match those examined so far within its native region in Southeast Asia.

B4. Natural enemies of Tachardina aurantiaca on Christmas Island and in Southeast Asia

Key outcomes: Few natural enemies of Tachardina aurantiaca occur on Christmas Island and they do not regulate its population size. No parasitoids of female T. aurantiaca were found, although Marietta leopardina parasitized a small fraction of males at some sites. Conversely, within its native range in Malaysia, T. aurantiaca is rare and patchily distributed, associated with diverse natural enemies, including at least five species of primary parasitoids, and suffers high parasitization rates – all attributes consistent with population regulation by natural enemies (Table B3, Fig. B3). Importantly, high parasitization rates in Malaysia occurred in the presence of honeydew-collecting ants, including the yellow crazy ant. Elsewhere, M. leopardina is a known hyperparasitoid of

3 primary parasitoids so a greater understanding of its relationship with any prospective biological control agent of T. aurantiaca on Christmas Island is needed. B5. Selection of prospective biological control agents of Tachardina aurantiaca on Christmas Island

Key outcome: Of the primary parasitoids known to attack yellow lac scale from our initial studies in Malaysia, Tachardiaephagus somervillei (Encyrtidae) (Fig. B4) is the most promising agent for introduction and release on Christmas Island. All Tachardiaephagus species have a narrow host range and appear to be family specialists, known only to attack the Kerriidae, the family to which the yellow lac scale belongs (Table B4). Our initial studies indicate that (a) T. somervillei attacks T. aurantiaca across 1900-km range in Peninsular Malaysia and Malaysian Borneo, (b) is the most abundant natural enemy of T. aurantiaca, (c) has a short life cycle compared to its host, (d) exhibits superparasitism (i.e., where multiple progeny emergence from a single host individual); (e) causes high rates of parasitism on T. aurantiaca in the presence of tending ants, including the yellow crazy ant, and (f) can be reared under laboratory conditions.

C. Research outcomes and the project goal

1. The key research outcomes (A1-A4, B1-B5) firmly support both the concept of indirect biological control and the safe introduction a host-specific biological control agent for Tachardina aurantiaca.

2. Prospects for island-wide suppression of YCA supercolonies by targeting this single species are uncertain but probably good.

3. The discovery of suitable agents for the control of soft scale insects (Coccidae) on Christmas Island provides the opportunity to target the entire assemblage of honeydew-producing scale insects, reducing uncertainty of achieving YCA supercolony suppression by the targeted control of T. aurantiaca alone.

4

B A

C D

Figure A1. Performance and foraging behavior of yellow crazy ants depend on sugar supply. Initially each laboratory colony contained two queens and 200 workers (n = 18 nests). A pair of colonies was assigned to one of nine treatments (0, 10, 20, 40, 80, 160, 320, 640, and 1280 ul) of 13% honey water that was delivered to each colony every 3-4 days for two months. YCAs had unlimited access to water and protein (thawed crickets). In the last week, novel objects (a small frame of bamboo skewers) were added to track YCA exploratory behaviour. (A) Per capita recruitment to sugar indicated that a smaller fraction of colony workers is needed to collect sugar with increasing sugar supply. Colony performance, as measured by (B) production of workers and males (P = 0.034) increased with sugar availability whereas (C) per capita death rate decreased with increasing sugar availability. (D) Per capita encounter rate of YCA with novel, non-food objects placed in the nest, a measure of foraging tempo, also increased with sugar supply.

5

A

B

Figure A2. Aggressive behaviours in the yellow crazy ant increased with sugar supply. Aggression assays between YCAs and the big-headed ant Pheidole megacephala were conducted by placing three YCAs and three Pheidole in a 6-cm diameter vial with fluon-coated sides for 10 minutes. The time to the first spray of formic acid by each YCA, the total number of sprays by each YCA, and the total number of dead P. megacephala dead after 10 minutes were recorded. YCA with access to more sugar sprayed P. megacephala with formic acid sooner (P = 0.006) and (A) more often, and (B) killed more P. megacephala in 3:3 interaction trials.

6

Figure A3. Stable isotope analyses (δ15N) of yellow crazy ants from four declining supercolonies showed that trophic position increased (i.e. YCA became more carnivorous) as population densities decreased over 18 months in 2010-2011 (Site 206, R2 = 0.75, P = 0.025; Site 318, R2 = 0.67, P = 0.025; Site 403, R2 = 0.58, P = 0.046, Site 582, R2 = 0.72, P = 0.033). Trophic position was calculated from mixing models that considered site- specific stable isotope ratios of known plants, herbivores, and carnivores. In this figure, a trophic position > 3 indicates stronger carnivory while < 3 indicates increasing herbivory. The datapoints are connected in temporal order for each site, beginning at the right-hand side.

7 A

50

B

)

1 -

40

30

20

10

YCA s 30 ants (no. counts trunk YCA 0

Figure A4. Banding trees effectively excluded YCA from honeydew resources in the forest canopy. (A) Tree bands were made by winding Gladwrap™ around the boles of all trees >5 cm DBH on a 40 m x 40 m plot. Bands blocked YCA traffic flow to and from the forest floor, resulting in “log jams” of downward moving ants, most replete with honeydew, above the bands. Throughout the experiment these stranded ants were returned to the forest floor by brushing them gently off each bole. Repeated application of Mr Sheen™ spray-on furniture polish to each band over the experiment greatly increased the effectiveness of the barriers. YCA trunk traffic was estimated by counts per 30 s in a 10 x 10 cm quadrat on each of 10 trees. (B) The experiment was a Before-after-control-impact (BACI) design on two forest plots in a YCA supercolony abutting the Winifred Beach Track. Treatment was an exclusion of YCA foragers from the forest canopy by banding trees or a control without bands. Response variables were YCA trunk traffic and YCA ground activity in the 20 x 20 m core of each plot, made at 3-4 day intervals (9 times before and after applying tree bands) during the dry season in 2012. Results were analyzed as a one-way repeated measures ANOVA, using ant exclusion from the forest canopy as the main factor and time as the repeated measures factor. In this design, the time x treatment interaction is the key term, with a significant treatment difference after, but not before treatment application. Bands effectively excluded YCA from the canopy, resulting in a precipitous decline and the virtual elimination of YCA traffic on tree boles after 4 weeks after the tree bands were in place (solid vertical line) (Treatment x Time interaction, F1,32 = 90.198, P = 0.000).

8 A

) 40 B

1 -

30

20

10

YCA ground counts (no. card30s YCA 0

Figure A5. Exclusion of the yellow crazy ant from the forest canopy resulted in a significant and rapid decline in YCA abundance on the forest floor. (A) YCA abundance on the ground was estimated by counts per 30 s on one quadrant on each of twelve 20 x 20 cm cards. (B) After tree bands were placed (solid vertical line), YCA abundance fell and diverged markedly from the control plot after two weeks (Treatment x Time interaction, F1,32 = 37.604, P = 0.000). YCA abundance declined ~3-fold compared to average pre-exclusion values, and was ~5-fold lower than on the control plot 4 weeks. If card counts are converted to YCA densities (using the regression y = 15.694x – 21.612; Abbott 2005, Insectes Soc. 52, 266), YCA densities on the forest floor fell from ~400 m-2 before to ~140 m-2 after exclusion from the canopy, and when the experiment was terminated were ~600 m- 2 on the control plot compared to ~100 m-2 on the exclusion plot. Plot disturbance when setting up the tree bands could have affected YCA activity on the ground but would be unlikely to explain the magnitude of change in YCA abundance after bands were in place.

9 100 90

in 2000 in 80 70 60

Tachardina Tachardina 50 40 30 20 10

% Contribution by by Contribution % 0 70 80 90 100 110 120 130 Site Honeydew Index in 2000

Figure A6. The yellow lac scale Tachardina aurantiaca is estimated to contribute a large fraction of the honeydew economy at forest sites with YCA supercolonies. At each site the estimated range in % contribution to SHI by T. aurantiaca is given, based on per capita parity in honeydew production by T. aurantiaca and coccoid scale insects (blue circles, mean = 70%, range 46-86%) or a 3 times greater per capita honeydew production by T. aurantiaca (red circles, mean 86.5%, range 72 – 95%). The Site Honeydew Index (SHI) is a relative estimate of the total volume of honeydew produced at a site. It is a compound measure that considers tree size, tree abundance, and the capacity of different species to host either Tachardina aurantiaca, or coccoid honeydew-producing species of soft scale insects. The basis of the Index is canopy surface area, generated for each tree using published allometric equations that predict crown dimensions from stem diameter (Poorter et al. 2006, Ecology 87:1289-1301). This estimate is then multiplied by the average abundance of scale insects per unit length of stem, using host species-specific data for both Tachardina and soft scales (data from Table 4.3 in Abbott 2004, PhD Thesis). The SHI is the sum of these products, divided by 10000 for convenience. Thus, the SHI is the sum of the contributions from individual trees for both Tachardina and soft scales, weighted by tree size (canopy surface area) and species identity. The SHI varied 1.6-fold (from 78 to 125) across the 10 forest plots with YCA supercolonies, suggesting that the capacity of forest stands to support YCA supercolonies varies considerably.

10

2002 2012

Figure A7. Estimated change in the Site Honeydew Index specifically for Tachardina aurantiaca on Inocarpus fagifer from 2000-2012 at eight 0.25 ha sites in former YCA supercolonies. Numbers on the y-axis are an index of Tachardina abundance and follow the methodology for calculation of the SHI in Figure A6 except is only applied to Inocarpus. The decline in SHI averaged 50% (range 11- 91%) of the 2000 value and was driven by the net decline in Inocarpus biomass across the sites resulting from Tachardina-induced mortality. These declines in Inocarpus were probably widespread in supercolonies that formed over a decade ago. If supercolonies re-form at these sites and there is no change in site capacity to host soft scale insects then the relative contribution of Tachardina will be lower. YCA supercolony suppression by the targeted control of T. aurantiaca is less certain under these circumstances. The extent of decline could be estimated from the Island-wide survey by determining the proportion of supercolonies that reform, and the proportion of time supercolonies occupied each site.

11

Table B1. Established collaborators and cooperators for the biological control project. Prof. Gullan and Dr. Cook provide taxonomic expertise on scale insects as well as morphological and genetic analyses for Tachardina aurantiaca on Christmas Island and in Southeast Asia. Dr. Hayat provides taxonomic expertise in the identification of encyrtid and aphelinid parasitoids associated with T. aurantiaca. FRIM provides benchspace and facilities for rearing T. aurantiaca and its natural enemies in Kepong, Selangor, Peninsular Malaysia. Ms Ong, a Masters student under supervision of Dr. Neumann, is conducting her research on the biology of T. aurantiaca and developing rearing techniques for it and one of its important natural enemies, Tachardiaephagus somervillei. Sarawak Forestry, Sabah Forestry, and the NUS provide laboratory space and field assistance. Dr. Pemberton, who has worked on the biological control of the invasive lobate lac scale in Florida, provides advice and was instrumental in establishing some of the cooperators listed below.

Location Collaborator Expertise

Australia ANU, Biological Sciences Prof. Penny Gullan Biology of scale insects University of Queensland, Dr. Lyn Cook Phylogeny and Biological Sciences genetics of scale insects India Aligarh Muslim University, Dr. Mohammed Hayat Microhymenopteran Aligarh, Uttar Pradesh taxonomy

Peninsular Malaysia Forestry Research Institute Dr. Laurence Kirton Entomology Malaysia (FRIM) Universiti Sains Malaysia Ms Ong Su Ping Entomology (M.Sc student) Sarawak Sarawak Forestry, Botanical Ms Lucy Chong Entomology Research Centre, Semenggoh Mr Het Bin Kaliang Entomology

Sabah Sabah Forestry Department, Dr. A. Chung Yaw Chyang Entomology Forest Research Centre, Dr. Chey Vun Khen Entomology Sepilok

Singapore National University of Dr. Rudolf Meier Entomology Singapore (NUS), Biological Dr. Hugh Tan Tiang Wah Entomology Sciences

United States USDA-ARS, Ft. Lauderdale FL Dr. Robert Pemberton Biological control

12 Table B2. Scale insects of Christmas Island. It is highly probably that all of these species, with broad host plant ranges and geographic distributions, are exotic to Christmas Island and introduced following human settlement. The yellow lac scale, Tachardina aurantiaca (Kerriidae) and Coccidae (soft scales) are the primary honeydew suppliers to the yellow crazy ant. Honeydew-producing scale insects in bold occur commonly in YCA supercolonies (G. Neumann, unpubl. results). Taxonomy and distribution from Ben-Dov et al. (2012, http://www.sel.barc.usda. gov/scalenet/scalenet.htm).

Honeydew Family and Species Common Name Distribution Producer

Kerriidae (lac scales) Tachardina aurantiaca (Cockerell) Yellow lac scale Oriental Yes *Paratachardina pseudolobata False lobate lac Oriental, Nearctic, Yes2 (Kondo & Gullan) scale Neotropical Coccidae (soft scales) Coccus celatus De Lotto Green coffee scale Afrotropical, Australasia, Yes Oriental *C. hesperidium Linnaeus Brown soft scale Cosmopolitan Yes *Milviscutulus mangiferae (Green) Mango shield scale Cosmopolitan Yes *Ceroplastes ceriferus (Fabricius) Indian wax scale Cosmopolitan Yes *C. destructor Newstead White wax scale Afrotropical, Australasia, Yes Oriental Saissetia oleae (Olivier) Black olive scale Pantropical Yes *S. coffeae (Walker) Hemispherical scale Cosmopolitan Yes **Parasaissetia nigra1 Nigra scale Cosmopolitan Yes **Pulvinaria urbicola Cockerell Urbicola soft scale Pantropical Yes **P. psidii1 Green shield scale Cosmopolitan Yes Monophlebidae (giant scales) Icerya purchasi (Maskell) Cottony cushion scale Cosmopolitan Yes

Cerococcidae (ornate pit scales) **Cerococcus indicus (Maskell) Spiny brown coccid Cosmopolitan Yes

Pseudococcidae (mealybugs) **Dysmicoccus finitimus Williams Asian coconut Australasia, Oriental Yes mealybug **Ferrisia virgata (Cockerell)1 Striped mealybug Cosmopolitan Yes

Diaspididae (armoured scales) Aspidiotus destructor (Signoret) Coconut scale Cosmopolitan No Pseudaulacaspis pentagona White peach scale Cosmopolitan No (Targioni Tozzetti) *Hemiberlesia palmae (Cockerell) Tropical palm scale Cosmopolitan No *Lindingaspis sp. -- -- No **Lepidosaphes sp.1 Oystershell scale -- No

*Record added by Abbott (2004, PhD Thesis, Monash University); **Record added by Neumann et al. (unpublished results); 1Tentative identification; 2P. pseudolobata produces honeydew but eject it instead of producing droplets that can be collected by ants (Howard 2010, Fla. Entomol. 93, 1).

13 Table B3. Natural enemy assemblages of the yellow lac scale Tachardina aurantiaca on Christmas Island and in Malaysia. + = present, -- = absent. For associates of T. aurantiaca, primary parasitoids oviposit on or in a host and develop within, ultimately killing the host. Hyperparasitoids seek out hosts with primary parasites, oviposit, and develop within the primary parasitoid. Predators feed externally and consume multiple scales. On Christmas Island, search focused on seven areas (see Figure B1) and examined over 11,000 females and 2000 males of the yellow lac scale. In Malaysia, targeted search for T. aurantiaca on known host plants (e.g., Pongamia pinnata) was frequently used to locate T. aurantiaca. Because the yellow lac scale is so rare across Malaysia, many fewer individuals were inspected for parasitization. To assess enemies, host plant twigs with aggregates of T. aurantiaca were first inspected, and then isolated so that emerging parasitoids could be collected for later identification. Individual females of T. aurantiaca were also isolated to collect emergent parasitoids and dissected to determine overall rates of parasitization. On Christmas Island, lac scale predators Eublemma sp. and ?Holcocera sp. were extremely rare.

Association with T. Christmas Species (Family) Malaysia aurantiaca Island Tachardiaephagus somervillei primary parasitoid -- + Mahdihassan (Encyrtidae) T. sarawakensis Hayat et al. primary parasitoid -- + (Encyrtidae) Coccophagus euxanthodes Hayat primary parasitoid -- + et al. (Aphelinidae)

C. tschirchii Mahdihassan primary parasitoid -- + (Aphelinidae)

Coccophagus sp. (Aphelinidae)1 primary parasitoid2 -- +

Promuscidea unfasciativentris hyperparasitoid -- + Girault (Aphelinidae)

Aprostocetus (syn. Tetrastichus) hyperparasitoid3 -- + purpureus Cameron (Eulophidae)1

Marietta leopardina Motschulsky 4 primary parasitoid + + (Aphelinidae)

Eublemma sp. (Noctuidae) predator + +

?Holcocera sp. (Blastobasidae) predator + +

1Tentative identification; 2Attack male T. aurantiaca only; 3primary parasitoid of many Coccidae, Diaspididae, Kerriidae, Margarodidae, and Pseudococcidae but known as a hyperparasitoid of C. tschirchii and Tachardiaephagus sp.; 4On Christmas Island and in Malaysia, Marietta leopardina is known only to attack male T. aurantiaca. It has never been observed emerging from female T. aurantiaca. In Southeast Asia, it is also a hyperparasitoid of primary parasitoids of a variety of scale insects.

14 Table B4. Records of known host families and genera for the primary parasitoid Tachardiaephagus (Encyrtidae). As a genus, Tachardiaephagus has an extremely broad geographic range. With the exception of one probably erroneous host record in Africa, all Tachardiaephagus species appear to be family specialists and restricted to the Kerriidae. For host genera, number of species recorded as hosts is in parentheses. Based on Noyes (2012, Universal Chalcidoidea Database, http://www.nhm.ac.uk/ research-curation/research/projects/chalcidoids/database/), except for records for T. somervillei and T. sarawakensis (Hayat et al. 2010. Oriental Insects 44, 23; R.W. Pemberton, pers. comm.; this study)

Species Distribution Recorded hosts

Tachardiaephagus India, Malaysia, Kerriidae somervillei Thailand Kerria spp. (4) Tachardina aurantiaca Tachardina sp. (1) T. sarawakensis Sarawak Kerriidae Tachardina aurantiaca T. tachardiae Brunei, China, India, Kerriidae Indonesia, Malaysia, Kerria spp. (6) Sri Lanka, Taiwan, Laccifer spp. (3) Vietnam, Azerbaijan Paratachardina lobata T. similis Afrotropical, Kerriidae South Africa Tachardina sp. (1) Coccidae ?Ceroplastes eucleae1 T. absonus Afrotropical, Kerriidae South Africa Tachardina spp. (2) T. communis Afrotropical, Kerriidae South Africa Tachardina spp. (5) T. gracilis Afrotropical, Kerriidae South Africa Tachardina sp. (1) 1This record is likely to be erroneous (see Prinsloo1983. Entomol. Mem. Dept. Agric. Rep. S. Afr. 60, 26).

15

Figure B1. Search methodologies for detecting native or endemic scale insects and natural enemies of the yellow lac scale T. aurantiaca on Christmas Island. Northwest Point is enlarged in the upper left-hand side of the figure. Search for native and endemic scale insects included (1) Timed searches (30 min) within a 50- m radius at each of 151 waypoints from the CINP island-wide survey. Searches were replicated three times in each dry season in 2010-2012. (2) Searches were conducted along the entire length of five tracks, a total of 21 km, twice in each of the dry seasons of 2011 and 2012 (Northwest Point Track, 4.7 km; Boulder Track, 6 km; Blowholes Road, 3.4 km; Martin Point Lookout to CINP boundary, 1.9 km; and, Dolly Beach Boardwalk, 1.8 km). Search effort totaled ~405 hours that did not include travel between sites. (3) Endemic plant species were a focal point of further search assuming that endemic scale insects would be more likely to be associated with them (Neumann et al. 2007, Proc. Hawaii. Entomol. Soc. 39, 39). Fifteen of the 18 species of endemic plant species were located and examined between 2010-2012. Ten rare endemic plants species from the CINP database were searched for scale insects at 125 locations. For each common endemic plant species, 30 haphazardly selected individuals were examined each year between 2010-2012. (4) CINP personnel also aided the search by being aware of scale insects during the island- wide survey in 2012. Search for natural enemies of T. aurantiaca involved seven focal areas on the island: The Dales (Hugh’s Dale to Sydney’s Dale - 4000 females, 2000 males), Martin Point to CINP boundary (1500 females), northwest of Central Area Workshop (200 females), Smith Point (1000 females), Dolly Beach Track (1000 females), Northwest Point track (1500 females), and the Circuit Road (2000 females). Each female and male was inspected under magnification to determine the presence of any parasitoid emergence holes.

16 Christmas Island

Figure B2. Christmas Island (red circle) is the only known area of introduction and invasion of the yellow lac scale Tachardina aurantiaca (Kerriidae). Populations of T. aurantiaca have now been located across a 1900-km east-west distribution in Sundaland (i.e., the part of Southeast Asian continental shelf that was exposed during the last ice age, encompassing Peninsular Malaysia, Borneo, Java, and Sumatra), its putative area of origin. Much of the search in Malaysia was targeted, using a highly suitable and widespread host plant, Pongamia pinnata, as a focus. In Malaysian Borneo, search effort totaled 27 days and 5 days were spent searching in Singapore. Sites with live aggregates of lac scale (yellow circles) in Peninsular Malaysia occurred on Penang Island, and in Klang (Selangor) and Kepong (Selangor). In Singapore, T. aurantiaca was found on the campus of the National University of Singapore. In Sarawak, live aggregates were discovered around Santubong and Kuching (two sites - Kampung Istana and Kampung Boyan). In Sabah, live aggregates were found in the Sandakan area and Sepilok. Dead aggregates (not shown above) were found in Melaka, peninsular Malaysia and in Temon, 190 km south of Kota Kinabalu, Sabah. Ants tended T. aurantiaca at all sites and collected honeydew (Penang – Crematogaster sp.; Klang – Oecophylla smaragdina; Kepong – Dolichoderus sp.; Kampung Istana – Anoplolepis gracilipes; Kampung Boyan – Oecophylla smaragdina; Singapore, Sandakan and Sepilok – Anoplolepis gracilipes). T. aurantiaca collected from 5 different sites and host plant species on Christmas Island are morphologically and genetically identical (based on cuticular morphology and mitochondrial cytochrome c oxidase subunit 1 (COI) and 28S ribosomal RNA sequences – P. Gullan and L. Cook, personal communications, 2012). Initial morphological and genetic studies of T. aurantiaca collected from sites in Sarawak indicate that they are morphologically and genetically identical to T. aurantiaca on Christmas Island (P. Gullan, personal communication, 2012).

17 A B

Figure B3. Parasitism of the yellow lac scale Tachardina aurantiaca in Malaysia. A. Parasitoid emergence holes of Tachardiaephagus sp. (Encyrtidae) in tests of an aggregate of old adult females of T. aurantiaca on Pongamia pinnata near Sandakan, Sabah, Malaysian Borneo. Yellow crazy ants (Anoplolepis gracilipes) tended T. aurantiaca at this site. B. Emergence holes of Coccophagus euxanthodes (Aphelinidae) in the test of a T. aurantiaca on Acacia auriculiformis at Klang, Selangor, Peninsular Malaysia. The smaller opening in the centre of the test is the anal pore through which honeydew is produced. Weaver ants (Oecophylla smaragdina) tended T. aurantiaca at Klang. Tachardiaephagus parasitized T. aurantiaca across all sites examined whereas Coccophagus species (C. tschirchii and C. euxanthodes) were uncommon and found only in Peninsular Malaysia. Parasitization rates by T. somervillei were generally high and most female Tachardina in aggregates had been killed (see A above). Incidence of parasitization based on emergence holes of T. somervillei in 5 aggregates of T. aurantiaca in Sarawak was 44.6 ± 9.6%; when corrected for parasitoid larvae that had not emerged, parasitization rate increased to 61.4 ± 7.7%. Superparasitism (i.e., where more than one progeny is produced per host individual) is frequent. The mean number of Tachardiaephagus somervillei emerging from each parasitized female T. aurantiaca was 2.1 (range 1- 4, N = 30) and up to 5 emergences occurred from each female in Peninsular Malaysia. In contrast, parasitism of T. aurantiaca on Christmas Island was nil; emergence holes were never seen in examination of over 11,000 females over two years (or over the past 15 years of observations of T. aurantiaca on the island). However, single exit holes were observed in some male T. aurantiaca at a few sites. In the laboratory the primary parasite Marietta leopardina emerged from these males. Parasitization rate was low, ranging from 0-10% (N= 558, 751, 696 males examined on three trees at Hugh’s Dale, Anderson’s Dale, and Sydney’s Dale), but M. leopardina is clearly not an effective biological control agent of female T. aurantiaca on the island. High densities of intact females occur where M. leopardina is present and, on Christmas Island, T. aurantiaca may be parthenogenetic (G. Neumann, unpublished results).

18 1 mm

Figure B4. Tachardiaephagus somervillei, a primary parasite of the yellow lac scale Tachardina aurantiaca in Malaysia and Singapore. In initial studies in 2011-2012, T. somervillei attacked T. aurantiaca across Peninsular Malaysia and Malaysian Borneo, is the most abundant natural enemy of T. aurantiaca, exhibited superparasitism (i.e., where multiple progeny emergence from a single host individual), and heavily parasitized T. aurantiaca in the presence of tending ants, including the yellow crazy ant. Ms Ong Su Ping, Dr. Neumann’s Master’s student, has also reared T. somervillei under laboratory conditions (drawing from Narayanan 1962. Pests of lac in India. Pp. 90-113 in Mukhopadhyay, B and Muthana, MS (eds) A Monograph on Lac. Indian Lac Res. Institut. Nancum, Ranchi, India).

19 PART A - Identifying and Analysing Project Management Risks

Division Parks Australia Project Title Research and development for indirect biological control of the yellow crazy ant (Anoplolepis gracilipes) on Christmas Island, Indian Ocean

Branch /Section Christmas Island National Park Project Goal Establish the feasibility of suppressing YCA supercolonies on Christmas Island by the safe introduction of a host-specific Date 21/11/2013 parasitoid of its lac scale mutualist, Tachardina aurantiaca

NOTE: In the Risk Assessment matrix below, the Consequence of The Risk can be judged in terms of:

1. Financial Consequences. For example, should the biocontrol program not proceed, the Consequence is recurrent funding for a business-as-usual approach to YCA management dependent on surveillance, monitoring and control by toxic bait. 2. Indirect Biodiversity Consequences, which arise not from the implementation of an action itself or from its failure, but from the knock-on effect of not implementing that action, or of it failing. For example, should the biocontrol program not proceed, the Consequence is continued negative impacts on elements of Island biodiversity, from either YCA or non-target impacts associated with the use of toxic bait. 3. Direct Biodiversity Consequences, which are a direct result of implementing an action. For example, off-target impacts of the biocontrol agent imported to the island would have Direct Biodiversity Consequences.

For Risk 1 (below), No further action on the Biocontrol Program, we have combined both Financial and Indirect Biodiversity considerations when judging the severity of the Consequences. For Risks 2 – 5, we judged severity for Direct Biodiversity Consequences only. In other words, would the biocontrol program lead to further decline in biodiversity and conservation values on the island?

20

1 2 3 4 5 6 7

Risk The Risk Ref/ID Source Consequence

(What Can Happen?) (How can this Happen?) (What will happen if the risk Risk

occurs?) Level

Likelihood Consequence

1 No further action is taken to a. Funding is insufficient to continue a. Business as usual – Possible Major High progress the biocontrol management of YCA research and development supercolonies continues to be project reactive. Formation of new supercolonies cannot be prevented, and can only be controlled after they form. Continuing impacts of supercolony formation on biodiversity, especially red crabs and robber crabs. Continuing financial investment in the human resources and infrastructure (toxic bait and helicopter) to detect, map and control supercolonies. Continuing non-target impacts of the toxin, especially on robber crabs

21 1 2 3 4 5 6 7

Risk The Risk Ref/ID Source Consequence

(What Can Happen?) (How can this Happen?) (What will happen if the risk Risk

occurs?) Level

Likelihood Consequence

b. Research does not support the concept of b. As above Unlikely Major Medium indirect biocontrol – YCA is not dependent on honeydew, or suitable biocontrol agents (BCAs) are not available

2 The imported BCA is not as Insufficient knowledge of the host specificity of the a. Other scale insects; In addition to Very Insignificant Minimal host specific as predicted BCA the target species Tachardina Unlikely aurantiaca, the BCA parasitizes other species of scale insects on the Island. All known species are non-native

b. Other arthropods, including Very Major Low endemic insects such as stick Unlikely insects and butterflies

22 1 2 3 4 5 6 7

Risk The Risk Ref/ID Source Consequence

(What Can Happen?) (How can this Happen?) (What will happen if the risk Risk

occurs?) Level

Likelihood Consequence

3 There is an endemic or Insufficient field survey effort The BCA will parasitize a native Very Major Low native scale insect in the species of scale insect Unlikely Family Kerriidae on Christmas Island that has not yet been discovered

4 The BCA does not achieve a. The climate on Christmas Island is wrong for a. The BCA will not establish Very Insignificant Minimal island-wide population the BCA Unlikely control of Tachardina aurantiaca to the low levels seen in Southeast Asia

b. The BCA is suited to parasitize an ecotype of b. The BCA will not establish well on Very Insignificant Minimal Tachardina from one part of its native range, Christmas Island Unlikely but a different ecotype occurs on Christmas Island

c. The BCA is a poor disperser, and cannot c. The BCA will only achieve local Very Insignificant Minimal move much beyond its initial point of release population control of Tachardina, Unlikely at points of release

23 1 2 3 4 5 6 7

Risk The Risk Ref/ID Source Consequence

(What Can Happen?) (How can this Happen?) (What will happen if the risk Risk

occurs?) Level

Likelihood Consequence

d. Tending of Tachardina by YCA reduces d. The incidence of parasitization on Unlikely Insignificant Minimal parasitization by BCA Tachardina by the BCA will be low

e. The BCA was infected with a hyperparasitoid e. The population densities of the Very Insignificant Minimal prior to importation BCA will never be high enough to Unlikely achieve control Tachardina

f. Marietta leopardina, already present on f. The population densities of the Possible Insignificant Low Christmas Island, acts as a hyperparasitoid for BCA will be too low to achieve the BCA once it is introduced control of the target

5 Targeting Tachardina The BCA for Tachardina is effective, but because Supercolonies continue to form Possible Insignificant Low aurantiaca alone is of local variation in host tree composition, there is across the island, requiring business insufficient to suppress YCA sufficient honeydew produced by soft scale as usual, surveillance, monitoring supercolonies, either locally, insects in some locations to permit the formation and control using toxic bait and islandwide, or both. of new ones. The net effect is the patchy considerable human resources suppression of YCA supercolonies

24

PART B – Risk Register & Corrective Strategies

1For justification of risk, see Executive Summary (ES)

Risk Short title Risk Level Justification for Likelihood & Options for Risk Mitigation Consequences1

1a Insufficient funding High Christmas Island National Park Sufficient funding for the Biocontrol Program Plan of Management Expert Working Group Final Report Reports and Publications by La Trobe University and Monash University

1b Negative research findings Medium ES A1-A4 None ES B1-B5

2a Non-target scale insects Minimal ES B4 Pre-importation host specificity trials conducted under laboratory conditions in Malaysia

2b Non-target other insects Low ES B4 Pre-importation host specificity trials conducted under laboratory conditions in Malaysia

3 Undiscovered native/endemic Low ES B2 None. Field surveys demonstrate to operational certainty scale insect there are no native or endemic scale insects on Christmas Island

4a Climate mismatching for the Minimal ES B3 & B5 Select donor populations of the BCA from areas that BCA match most closely the local climate on Christmas Island

25 Risk Short title Risk Level Justification for Likelihood & Options for Risk Mitigation Consequences1

4b Ecotype mismatching for Minimal ES B3 Select donor populations of the BCA from Tachardina Tachardina hosts that match most closely the local type on Christmas Island

4c Lack of dispersal by the BCA Minimal ES B3 Create “propagule pressure” by releasing the BCA at multiple sites and multiple times across Christmas Island

4d Scale tending by YCA Minimal ES B4 & B5 Select donor populations of the BCA from sites where Tachardina is tended by high densities of mutualistic ants, especially Anoplolepis gracilipes

4e Hyperparasitism of the BCA Minimal ES B4 Use standard laboratory rearing techniques to ensure the before introduction to the BCA is free of any hyperparasitoids prior to its importation island to Christmas Island

4f Marietta acts as a Low ES B4 None. However, there is no evidence of that Marietta hyperparasitoid of BCA after attacks female Tachardina, either as a primary parasitioid introduction to the island or in order to attack the larvae of other primary parasitoids inside female Tachardina.

26 Risk Short title Risk Level Justification for Likelihood & Options for Risk Mitigation Consequences1

5 Biocontrol of Tachardina Low ES A4 Expand program capacity to target the entire assemblage insufficient to suppress YCA of honeydew-producing scale insects, reducing supercolonies uncertainty of achieving suppression of YCA supercolonies by targeting T. aurantiaca alone. Develop and implement a biocontrol program for soft scale insects on Christmas Island to complement the program under development for Tachardina. Coccophagus ceroplastae (Aphelinidae) and Encyrtus infelix (Encyrtidae), both parasitoids of coccid scale insects, are already present on Christmas Island. Both attack a wide variety of soft scales, including three that are important in YCA supercolonies (Table B2 - Coccus hesperidium, Saissetia oleae, S. coffeae). The difficult issues of foreign exploration, host- specificity testing and navigating regulatory frameworks have been obviated by the presence of these parasitoids on the island. Experience with the efficacy of these agents in dealing with the outbreak of Pulvinaria urbicola (Neumann et al. 2011, report to CINP) suggests that current lack of control of soft scales in supercolonies is a result of dispersal limitation. Rearing and redistribution of C. ceroplastae did not require a referral under the EPBC Act.

27