Part 2 Landscape protection and restoration

13. Biological control of cat’s claw creeper (Macfadyena unguis-cati)

Project dates Field-release and monitoring Field-release and monitoring September 2002 – June 2014 We mass-rear and field-release two No further field-releases of the leaf- biological control agents, the leaf- sucking tingid (C. visenda) were made Project leader sucking tingid (Carvalhotingis visenda) after June 2010. Field-release of the Dr K. Dhileepan and leaf-tying moth (Hypocosmia leaf-tying moth (H. pyrochroma) was Ecosciences Precinct pyrochroma), in partnership with continued until October 2010. Over three Tel: (07) 3255 4449 community groups. We use a simple years, 1272 adult moths, 77 750 mature Email: [email protected] and cost-effective method to mass-rear larvae and 837 pupae have been released the leaf-tying moth by replacing potted across 36 sites in Queensland and New Other staff in 2010–11 plants with field-collected cut foliage South Wales. Field establishment status to allow greater numbers of insects of the moth was not monitored due to Di Taylor and Mariano Treviño to be released in the field. After field- relocation to the Ecosciences Precinct. Objective release we conduct recovery surveys to determine the field establishment status Funding in 2010–11 Achieve biological control of cat’s claw of C. visenda and H. pyrochroma. At • Land Protection Fund ($148 000) creeper using introduced insect species. all release sites, we spend 20 minutes visually examining cat’s claw creeper • Queensland Government (Blueprint Rationale plants and recording the incidence and for the Bush) Cat’s claw creeper (Macfadyena unguis- abundance of C. visenda eggs, nymphs Collaborators cati), an invasive liana native of Central and adults and H. pyrochroma larvae. and South America, is a major weed • Stefan Neser and Anthony King in coastal Queensland and New South Progress in 2010–11 [Agricultural Research Council—Plant Wales, where it poses a significant Protection Research Institute (ARC- threat to biodiversity in riparian and Host-specificity tests PPRI), South Africa] rainforest communities. The plant is We completed host-specificity tests • Dr Tanya Scharaschkin (Queensland a structural parasite and produces of the leaf-mining buprestid beetle University of Technology, Science and stolons and subterranean root tubers. (H. jureceki) in quarantine at the Engineering Faculty) Biological control appears the most Ecosciences Precinct in May 2011. suitable management option for this These tests support previous studies • Local government and community weed. Management objectives focus from South Africa indicating that the groups across south-eastern and on reducing the rate of shoot growth beetle is highly host-specific and does central Queensland to limit the weed’s ability to climb and not pose risk to any non-target plants smother native vegetation, as well as in Australia. Minor exploratory adult More information reducing tuber biomass to minimise the feeding occurred on eight non-target tuber bank. species and oviposition on one non- Key publications target species, but larval development Dhileepan, K, Treviño, M, Bayliss, D, Methods occurred only on M. unguis-cati. Saunders, M, Shortus, M, McCarthy, J, Observations indicate that this is a Snow, EL et al. 2010, ‘Introduction and Host-specificity tests highly damaging insect with two establishment of Carvalhotingis visenda Host-specificity testing is conducted destructive life stages: larvae mine (Hemiptera: Tingidae) as a biological using potted test plants in a within the leaves and adults chew holes control agent for cat’s claw creeper temperature-controlled (22 °C to 27 °C) into leaves. Under laboratory conditions, Macfadyena unguis-cati (Bignoniaceae) quarantine insectary. We evaluate the high populations can completely in Australia’, Biological Control potential host range of the leaf-mining defoliate cat’s claw creeper plants. 55(1): 58–62. buprestid beetle (Hylaeogena jureceki) A short generation time, long- Dhileepan, K, Bayliss, D & Treviño, M on the basis of larval survival and living adults and predator-evading 2010, ‘Thermal tolerance and potential development, adult feeding and survival, characteristics suggest that rapid distribution of Carvalhotingis visenda and oviposition preference using choice population growth is likely in the (Hemiptera: Tingidae), a biological and no-choice tests involving 38 plant field. We have applied to the relevant control agent for cat’s claw creeper, species in 12 families. regulatory authorities to release this Macfadyena unguis-cati (Bignoniaceae)’, agent in Australia. Bulletin of Entomological Research 100(2): 159–66.

Part 2 Landscape protection and restoration 29 Rafter, MA, Wilson, AJ, Wilmot Raghu, S, Dhileepan, K & Treviño, M Senaratne, KAD & Dhileepan, K 2008, 2006, ‘Response of an invasive liana to ‘Climatic-requirements models of cat’s simulated herbivory: implications for its claw creeper Macfadyena unguis-cati biological control’, Acta Oecologica (Bignoniaceae) to prioritise areas for 29(3): 335–45. exploration and release of biological Raghu, S, Wilson, JR & Dhileepan, K control agents’, Biological Control 2006, ‘Refining the process of agent 44(2): 169–79. selection through understanding Conrad, KA & Dhileepan, K 2007, ‘Pre- plant demography and plant response release evaluation of the efficacy of the to herbivory’, Australian Journal of leaf-sucking bug Carvalhotingis visenda Entomology 45(4): 308–16. (Heteroptera: Tingidae) as a biological Raghu, S & Dhileepan, K 2005, ‘The control agent for cat’s claw creeper value of simulating herbivory in Macfadyena unguis-cati (Bignoniaceae)’, selecting effective weed biological Biocontrol Science and Technology control agents’, Biological Control 17(3): 303–11. 34(3): 265–73. Dhileepan, K, Snow, EL, Rafter, MA, Dhileepan, K, Treviño, M, Donnelly, GP Treviño, M, McCarthy, J & & Raghu, S 2005, ‘Risk to non-target Senaratne, KADW 2007, ‘The leaf-tying plants from Charidotis auroguttata moth Hypocosmia pyrochroma (Lep., (Chrysomelidae: Coleoptera), a potential Pyralidae), a host-specific biological biocontrol agent for cat’s claw creeper control agent for cat’s claw creeper Macfadyena unguis-cati (Bignoniaceae) Macfadyena unguis-cati (Bignoniaceae) in Australia’, Biological Control in Australia’, Journal of Applied 32(3): 450–60. Entomology 131(8): 564–8. For further information on this research Dhileepan, K, Treviño, M & Snow, EL project and access to key publications, 2007, ‘Specificity of Carvalhotingis visit the invasive plant and animal science visenda (Hemiptera: Tingidae) as pages on the Biosecurity Queensland a biological control agent for cat’s website at www.biosecurity.qld.gov.au claw creeper Macfadyena unguis-cati (Bignoniaceae) in Australia’, Biological Control 41(2): 283–90. Raghu, S, Dhileepan, K & Scanlan, JC 2007, ‘Predicting risk and benefit a priori in biological control of invasive plant species: a systems modelling approach’, Ecological Modelling 208(2–4): 247–62.

30 Technical highlights: research projects 2010–11 14. Biological control of Madeira vine (Anredera cordifolia)

Project dates We then seek approval from the relevant Mass-rearing of the insect commenced regulatory authorities to release any immediately within non-quarantine June 2007 – June 2013 suitable agents. Approved agents are then glasshouses at the Ecosciences Precinct. Project leader mass-reared for distribution to climatically The insect was reared most satisfactorily favourable areas. Following release, in these facilities and produced sufficient Dr Bill Palmer we monitor establishment progress and numbers to begin field releases. Ecosciences Precinct evaluate any effects of the agents. Tel: (07) 3255 4469 By June 2011, we had released over 2000 Email: [email protected] Progress in 2010–11 beetles, primarily in the western suburbs of Brisbane. Due to the approaching Other staff in 2010–11 Host-specificity testing of the Madeira winter, this was not an optimal time for vine leaf beetle (Plectonycha correntina) insect survival. However, these releases Wilmot Senaratne, Liz Snow and concluded in late 2009 and showed that will help us gain a better understanding Peter Jones the beetle could use only Madeira vine of the insect’s overwintering capabilities. as a host. The release of P. correntina Objective for the biological control of Madeira We have also provided adult beetles to the New South Wales Government Achieve biological control of Madeira vine was therefore recommended Department of Primary Industries. This vine by introducing and releasing exotic in a report submitted to Australian is to establish a mass-rearing program at insect species or pathogens. regulatory authorities in December 2009. The case for this insect was the the Grafton research station to provide Rationale first to be processed through Biosecurity insects for release in New South Wales. Madeira vine (Anredera cordifolia) Australia’s new protocols for biological Dr Neser made a further exploratory trip is a South American plant that has control agents. to South America in the hope of finding become an increasingly important In early 2011, both regulatory populations of the leaf beetle Phenrica sp. environmental weed in eastern authorities approved the submissions, or other potential biocontrol agents, Australia. This vigorous perennial and the insect was released from but no insect species was forthcoming. climber or scrambling shrub forms quarantine in April 2011. It would be highly desirable to begin dense mats that cover trees and shrubs research on a second agent to supplement and it is now a problem weed in P. correntina eventually. rainforests, riparian lands, bushland remnants and conservation areas. It is the only naturalised plant in the family Basellaceae in Australia, so there is a good chance that biological control agents found in its native range would be sufficiently host-specific for safe release. However, one exotic species from this family, Ceylon spinach (Basella alba), is grown in gardens in south- eastern Queensland. South African scientists led by Dr Stefan Neser have identified some promising agents, which were made available to this project.

Methods Surveys for suitable biological control agents of Madeira vine are conducted Photo 14.1 Experimentalist Liz Snow releases P. correntina on a Brookfield property with very in Argentina and Brazil by Dr Stefan heavy Madeira vine infestation Neser from the ARC‑PPRI, South Africa. We import those insects considered suitable, mainly as a result of host testing undertaken in South Africa, into quarantine facilities in Australia for final host-specificity testing and biology studies. We also develop climate- matching models for prospective agents.

Part 2 Landscape protection and restoration 31 Funding in 2010–11 More information

• Land Protection Fund ($193 000) Key publications • Queensland Government Cagnotti, C, McKay, F & Gandolfo, D Collaborators 2007, ‘Biology and host specificity of Plectonycha correntina Lacordaire • Dr Stefan Neser and Liamé van der (Chrysomelidae), a candidate for the Westhuizen (ARC‑PPRI, South Africa) biological control of Anredera cordifolia • New South Wales Government (Tenore) Steenis (Basellaceae)’, African Department of Primary Industries Entomology 15(2): 300–09. Vivian-Smith, G, Lawson, BE, Turnbull, I & Downey, PO 2007, ‘The biology of Australian weeds 46. Anredera cordifolia (Ten.) Steenis’, Plant Protection Quarterly 22(1): 2–10. van der Westhuizen, L 2006, The evaluation of Phenrica sp. 2 (Coleoptera: Chrysomelidae: Alticinae), as a possible biological control agent for Madeira vine, Andredera cordifolia (Ten.) Steenis in South Africa, MSc thesis, Department of Zoology and Entomology, Rhodes University, South Africa. For further information on this research project and access to key publications, visit the invasive plant and animal science pages on the Biosecurity Queensland website at www.biosecurity.qld.gov.au

32 Technical highlights: research projects 2010–11 15. Biological control of lantana (Lantana camara)

Project dates Ongoing

Project leaders Michael Day Ecosciences Precinct Tel: (07) 3255 4453 Email: [email protected] Catherine Lockett Ecosciences Precinct (until December 2010) Tel: (07) 3255 4456 Email: [email protected] Cairns

Other staff in 2010–11 Natasha Riding and Kelli Pukallus Townsville

Objective O. camarae established

Import, evaluate host-specificity of, Mackay O. camarae release sites mass-rear, field-release and monitor major towns biological control agents for lantana.

Rationale Rockhampton Lantana (Lantana camara) is native to tropical America and was first Bundaberg introduced into Australia in the mid- 1800s. It has since become a major Gympie weed of agricultural and natural ecosystems. In grazing lands, lantana Brisbane dominates preferred pasture species (thereby decreasing productivity) and also interferes with mustering. Some Figure 15.1 Current distribution of O. camarae in Queensland varieties are toxic to livestock. It is estimated that this weed costs the We then determine the host-specificity Here, apart from having to cope with grazing industry over $100 million of imported organisms. Any agents the cooler conditions, the fly has to each year in lost production and control approved for field release are mass- compete with other biocontrol agents, expenses. Lantana can become the reared and released in appropriate areas particularly Calycomyza lantanae and dominant understorey species in natural with the help of Biosecurity Queensland, Uroplata girardi. ecosystems, blocking succession and DERM and local government field staff. decreasing biodiversity. Lantana is a To comply with Australian regulatory Class 3 declared weed in Queensland Progress in 2010–11 authorities, we contracted ARC‑PPRI to and has been the target of biocontrol conduct host-specificity testing of the The lantana herringbone leaf-mining programs since 1914. Introducing new lantana budmite (Aceria lantanae). The fly (Ophiomyia camarae) continues to be and more effective biocontrol agents studies found the budmite was sufficiently mass-reared at TWRC for field-release. further enhances control of lantana and specific to release. We have now gained However, field-releases in south-eastern reduces dependency on chemicals and approval from Biosecurity Australia to Queensland have stopped, since the fly other control methods. release the budmite. Approval from the appears to be more suited to tropical Australian Government Department Methods regions. Leaf mines have been found at of Sustainability, Environment, Water, more than 200 sites from Cooktown to Population and Communities is pending. We contract entomologists and Shoalwater Bay, with insects moving up to pathologists in Mexico, South Africa and 50 km away from some release sites. There Europe to locate and study the biology, has been substantial defoliation of lantana phenotype preference and preliminary bushes around Cooktown and Ayr. In host-specificity of potential biocontrol south-eastern Queensland, the fly has been agents prior to their introduction into found in low numbers at only a few sites. quarantine in Australia.

Part 2 Landscape protection and restoration 33 More information Key publications Day, MD & Nahrung, HF 2010, ‘Preference and performance of Aconophora compressa Walker (Hemiptera: Membracidae) on different lantana phenotypes in Australia’, Australian Journal of Entomology 49(4): 363–8. Day, MD, Riding, N & Chamberlain, A 2009, ‘Biology and host range of Ophiomyia camarae Spencer (Diptera: Agromyzidae), a potential biocontrol agent for Lantana spp. (Verbenaceae) in Australia’, Biocontrol Science and Technology 19 (6): 627–37.

Photo 15.1 Galls on lantana caused by A. lantanae Day, MD & Zalucki, MP 2009, ‘Lantana camara Linn. (Verbenaceae)’, in R Muniappan, GVP Reddy & A Raman We have imported A. lantanae into the Funding in 2010–11 (eds), Biological control of tropical weeds quarantine at the Ecosciences Precinct and using arthropods, Cambridge University Land Protection Fund ($129 000) are currently testing it against the main Press, Cambridge, pp. 211–46. lantana phenotypes in Australia. It readily causes gall formation on the red-flowering Collaborators Zalucki, MP, Day, MD & Playford, J 2007, and pink-edged red-flowering types, • ARC-PPRI, South Africa ‘Will biological control of Lantana camara and causes some gall formation on the ever succeed? Patterns, processes & • CABI Europe–UK, United Kingdom orange-flowering type. However, gall prospects’, Biological Control 42(3): 251–61. • Centre for Origin Research, United States formation has not been observed on the Day, MD, Broughton, S & Hannan- common pink-flowering type. Field- • CSIRO Plant Industry Jones, MA 2003, ‘Current distribution releases are expected to commence in • CSIRO Ecosystem Sciences and status of Lantana camara and its the summer of 2011–12. biological control agents in Australia, • DERM We contracted CABI Europe–UK to with recommendations for further • Office of Environment and Heritage, study the biology, biotype preference biocontrol introductions into other New South Wales and host-specificity of the pathogen countries’, Biocontrol News and Puccinia lantanae. Investigations found • New South Wales Government Information 24(3): 63N–76N. Department of Primary Industries some infection on the native Verbena Day, MD, Wiley, CJ, Playford, J & officinalis var. africana (previously • The University of Queensland Zalucki, MP 2003, Lantana: current reported as V. gaudichaudii) and • Local governments in Queensland and management status and future prospects, V. of ficinalis var. gaudichaudii as well New South Wales Australian Centre for International as the weed Phyla canescens. A report Agricultural Research, Canberra, 128 pp. has been received and a decision about conducting further testing is pending. For further information on this research project and access to key publications, visit the invasive plant and animal science pages on the Biosecurity Queensland website at www.biosecurity.qld.gov.au

34 Technical highlights: research projects 2010–11 16. Biological control of mikania vine (Mikania micrantha) in Papua New Guinea and Fiji

Project dates Better control of the weed in We field-release agents throughout areas neighbouring countries such as Papua of Fiji and Papua New Guinea where July 2006 – June 2012 New Guinea and Fiji will in turn mikania vine is a problem. Provincial Project leader reduce the risk of further spread into staff assist in the release of agents as part Queensland. A greater understanding of of their training in biocontrol activities. Michael Day mikania vine and its biocontrol agents We also set up programs for monitoring Ecosciences Precinct will also boost the state’s capacity plant density and spread as well as for Tel: (07) 3255 4453 to respond to an incursion if the agent establishment, population increase, Email: [email protected] eradication program is unsuccessful. spread and impact on mikania vine.

Objectives Methods Progress in 2010–11 Introduce and establish biocontrol agents Suitable agents are selected based on Mass-rearing and field-release of for mikania vine in Fiji and Papua New results of host-specificity testing and P. spegazzinii in both Fiji and Papua Guinea to: field observations in other countries. New Guinea continued throughout We send information on the agents’ 2010–11. In Papua New Guinea, the • reduce the impact of the weed to life histories and host ranges to our rust was released at nearly 300 sites in small-block holders and plantation collaborators in Fiji and Papua New 12 provinces in which the presence of owners in areas where it is a problem Guinea, and request import permits. mikania vine is confirmed. Pustules • reduce the seed load and thus the For the mikania butterflies (A. anteas have now persisted on field plants at possibility of further spread into and A. thalia pyrrha), additional host more than 120 sites in 10 provinces and northern Australia testing will be conducted in Fiji prior some pustules have been found more • establish successful biocontrol to field release. For the mikania rust than 30 km from the point of release methods for use in northern Australia (P. spegazzinii), CABI Europe–UK after 12 months. It is still too early, if required conducts additional host testing. We or the sites are too distant, to confirm submit reports on the host testing of the establishment at many of the other sites. • promote biocontrol as a safe and agents to quarantine authorities in Fiji successful weed-control method and Papua New Guinea. On approval, Field monitoring of several release sites • train scientists in Fiji and Papua New suitable agents are reared for release in around the research station at Kerevat Guinea in biocontrol methods. both countries. found that the rust is having a severe impact on mikania vine. At one plot the Rationale rust suppressed the growth of mikania Mikania vine (Mikania micrantha) is vine, which allowed other plants to smother native to tropical America and is now the plant, reducing its cover substantially. a major weed throughout the South Pacific and South-East Asia. The plant is a perennial vine that grows extremely rapidly, about 1 m month–1, smothering crops and plantation trees. In Queensland, mikania vine is currently confined to the wet tropics region, where it has the potential to significantly affect the sugar, horticultural, beef and tourist industries and to spread throughout northern Australia. Mikania vine is a Class 1 declared weed in Queensland and is the target of a national cost-share eradication program. Biological control of mikania vine in the South Pacific was first attempted in the 1970s. However, the agent failed to establish. This project aims to introduce two butterfly species (Actinote anteas and A. thalia pyrrha) Photo 16.1 Jenitha Fidelis from the Cocoa and Coconut Research Institute, Papua New Guinea, from Indonesia and the mikania rust showing rust-infected mikania to villagers in East New Britain Province, Papua New Guinea (Puccinia spegazzinii) into both Fiji and Papua New Guinea.

Part 2 Landscape protection and restoration 35 Funding in 2010–11 Australian Centre for International Agricultural Research ($70 000)

Collaborators • Australian Centre for International Agricultural Research • Secretariat of the Pacific Community • Ministry of Primary Industries, Fiji • National Agricultural Research Institute, Papua New Guinea • Cocoa and Coconut Research Institute, Papua New Guinea • Papua New Guinea Oil Palm Research Photo 16.2 Rust pustules on mikania vine in East New Britain Province, Papua New Guinea Association • CABI Europe–UK, United Kingdom P. spegazzinii established Lorengau P. spegazzinii released • Roch Desmier de Chenon, Consultant, but not checked Kavieng Indonesia Vanimo P. spegazzinii not released Wewak More information Rabaul Key publications Tabubil Madang Kimbe Ellison, C & Day, M 2011, ‘Current status Aropa of releases of Puccinia spegazzinii for Lae Mikania micrantha control’, Biocontrol News and Information 32(1): 1N–2N. Orapa, W, Day, M & Ellison, C 2008, Popondetta ‘New efforts at biological control of Port Moresby Mikania micrantha H.B.K. (Asteraceae) in Papua New Guinea and Fiji’, in Alotau 0 50 100 200 300 400 500 Proceedings of the Australia and New km Zealand IOBC Biocontrol Conference, Sydney, p. 45. Figure 16.1 Current distribution of mikania vine and P. spegazzinii in Papua New Guinea Pene, S, Orapa, W & Day, M 2007, ‘First9 fungal pathogen to be utilized 50 50 average number of leaves infected for weed biocontrol in Fiji and Papua

average number of petioles infected New Guinea’, Biocontrol News and

average number of stems infected Information 28(3): 55N–56N.

) 40 40 –2 mikania vine cover For further information on this research project and access to key publications, 30 30 visit the invasive plant and animal science pages on the Biosecurity Queensland website at www.biosecurity.qld.gov.au 20 20 Mikania vine cover (%) cover vine Mikania

10 10 Rust abundance (infected plant parts m plant (infected abundance Rust

0 0 Apr. June Aug. Oct. Dec. Feb. Apr. June Aug. Oct. Dec. Feb. Apr. June 2009 2010 2011 Date

Figure 16.2 Effect of P. spegazzinii on mikania vine cover in an experimental plot at Kerevat, Papua New Guinea

36 Technical highlights: research projects 2010–11 17. Weed eradication and containment: feasibility and program evaluation

Project dates to slow its spread (partial containment) resulting from abiotic dispersal, such as should eradication prove impossible. wind and water, rather than relying on July 2003 – June 2013 As some weeds may be more readily large generic dispersal buffers. Project leader contained than others, it is important to better define the scope for partial Assessing feasibility of Dr Dane Panetta containment. containment Ecosciences Precinct The reduction of seed production in Tel: (07) 3255 4472 Ongoing eradication and containment weed infestations is likely to play Email: [email protected] feasibility work should contribute to management decisions. Case-study data an important part in containment. Other staff in 2010–11 is needed to determine to what degree However, this is the case for virtually management objectives have been all weeds and hence not a particularly Simon Brooks and Shane Campbell achieved and to assess progress towards discriminating feature for assessing the eradication or containment. feasibility of containment. Therefore, Objectives our research has focused on the fates • Provide a scientifically based Methods of seeds that are produced. That is, while management interventions can rationale for decisions about the We collate data on eradication resources reduce propagule production (fecundity eradication and containment of weed and progress for each infestation of control) in source populations, weed incursions. clidemia (Clidemia hirta), limnocharis spread can also be reduced through • Refine eradication methods by using (Limnocharis flava), miconia actions that interfere with dispersal ecological information. (Miconia calvescens, M. nervosa and and establishment, and/or lead to M. racemosa), mikania vine (Mikania • Monitor selected eradication programs the detection and control of new micrantha)—under the National Four and document associated costs. infestations. Undetected new infestations Tropical Weeds Eradication Program— may reproduce and give rise to further • Develop criteria for assessing the and Siam weed (Chromolaena odorata) dispersal (shown as the dotted line in progress of eradication. in Queensland. Data includes method of Figure 17.1) . • Develop procedures for assessing the detection, discovery over time, trends in feasibility of containment. infested areas, population decline and Three case studies have been used in time since last detection. initial qualitative assessments of the feasibility of containment (Figure 17.2). Rationale Research on assessing the feasibility The distance of dispersal affects feasibility Early intervention is the most cost- of containment has adopted both of containment, but the ability to effective means of preventing weed qualitative and quantitative approaches. incursions from rapidly expanding. predict where seeds are deposited is also Strategies to achieve this aim range Progress in 2010–11 important, since this influences both the from eradication (where the objective potential for detection of new infestations is to drive the incursion to extinction) ‘Four tropical weeds’ database and the effort required to do so. Branched broomrape (Orobanche ramosa, a parasitic to containment (which may vary from We are currently converting the annual weed targeted for eradication absolute to degrees of slowing its spread). eradication progress reporting units in South Australia) is considered most from a buffered infestation area In previous research we have developed containable, since its dispersal is mainly to fixed management areas of 1 ha measures for evaluating eradication human-mediated and can be traced. (100 m × 100 m). This transition has progress with regard to the delimitation Parthenium is similar, but its seeds can been accompanied by more intense (determining the extent of the incursion) be dispersed widely by floodwaters, so recording of plant presence and absence and extirpation (local extinction) new infestations could be more difficult and will enable more consistent and criteria. We also have developed to find. The feasibility of containment of accurate reporting from year to year dynamic models that provide estimates miconia is considered lowest, since it is over fixed areas. The transition will of eradication program duration (and dispersed primarily by birds in relatively also enable portions of infestations total program costs when economic data unpredictable directions. is available). to progress to a monitoring stage and improve the progression factors for large Quantitative assessment of the feasibility Research conducted worldwide over infestations, including the melastome of containment is based on a model that the past decade has shown that the species known only at single locations. includes the area that must be searched, circumstances under which weed A finer scale of recording and reporting the effort required to detect a weed eradication can be achieved are highly can also increase the discovery rate of within this area and the potential for constrained, suggesting that eradication management areas in the short term colonisation and establishment in sites may not be widely applicable as a and increase the rate of reversion where foci of infestation are unlikely to weed-invasion management strategy. (from monitoring to control status). be detected. However, if a weed is sufficiently serious Eventually, this new spatial regime will to consider eradication there may often enable field crews to target searching be a strong justification for attempting to areas potentially containing plants

Part 2 Landscape protection and restoration 37 Fox, JC, Buckley, YM, Panetta, FD, Source population Bourgoin, J & Pullar, D 2009, ‘Surveillance protocols for management of invasive plants: modelling Chilean Propagule production needle grass (Nassella neesiana) in Australia’, Diversity and Distributions 15(4): 577–89. Dispersal Long, RL, Steadman, KJ, Panetta, FD & Adkins, SW 2009, ‘Soil type does Propagule pressure not affect seed ageing when soil water Habitat suitability potential and temperature are controlled’, Plant and Soil 320(1–2): 131–40. Panetta, FD 2009, ‘Weed eradication: Establishment Detection an economic perspective’, Invasive Plant Science and Management 2(4): 360–8. Figure 17.1 Schematic diagram showing the dispersal–establishment–detection cycle Brooks, SJ, Panetta, FD & Galway, KE 2008, ‘Progress towards the eradication of mikania vine (Mikania micrantha) and limnocharis (Limnocharis flava) branched broomrape in northern Australia’, Invasive Plant Science and Management 1(3): 296–303. 1 2 Long, RL, Panetta, FD, Steadman, KJ, Probert, R, Bekker, R, Brooks, SJ & parthenium Adkins, SW 2008, ‘Seed persistence in the field may be predicted by laboratory- controlled aging’, Weed Science 56(4): 523–28. 3 miconia 4 Panetta, FD 2007, ‘Evaluation of weed eradication programs: containment and extirpation’, Diversity and Distributions Predictability of direction of dispersal of direction of Predictability Distance of dispersal 13(1): 33–41. Panetta, FD & Lawes, R 2007, ‘Evaluation Figure 17.2 Feasibility of weed containment as influenced by combinations of distance and of the Australian branched broomrape predictability of direction of dispersal events, with basic scenarios ranked 1–4 in order of decreasing feasibility of containment, and showing relative positions of case study species along (Orobanche ramosa) eradication with potential variability in each dimension program’, Weed Science 55(6): 644–51. More information Regan, TJ, McCarthy, MA, Baxter, PWJ, Funding in 2010–11 Panetta, FD & Possingham, HP 2006, Queensland Government Key publications ‘Optimal eradication: when to stop Panetta, FD, Csurhes, SM, Markula, A looking for an invasive plant’, Ecology Collaborators & Hannan-Jones, MA 2011, ‘Predicting Letters 9(7): 759–66. • Oscar Cacho (University of New the cost of eradication for 41 Class 1 Panetta, FD & Lawes, R 2005, ‘Evaluation England) declared weeds in Queensland’, Plant of weed eradication programs: the Protection Quarterly 26(2): 42–6. delimitation of extent’, Diversity and • Biosecurity Queensland (South Distributions 11(5): 435–42. Johnstone) and local government Hester, SM, Brooks, SJ, Cacho, OJ & staff—provided data for eradication Panetta, FD 2010, ‘Applying a simulation For further information on this research case studies model to the management of an project and access to key publications, infestation of Miconia calvescens in the visit the invasive plant and animal science wet tropics of Australia’, Weed Research pages on the Biosecurity Queensland 50(3): 269–79. website at www.biosecurity.qld.gov.au Brooks, SJ, Panetta, FD & Sydes, TA 2009, ‘Progress towards the eradication of three melastome shrub species from northern Australian rainforests’, Plant Protection Quarterly 24(2): 71–8.

38 Technical highlights: research projects 2010–11 18. Ecology and control of national weed eradication targets

Project dates Methods • Some infestations in remote or rugged country cannot be treated July 2008 – June 2013 Specific questions investigated by this with high-volume foliar herbicide project and the trials to address them are applications and there has been Project leader outlined below. a reliance on manual control. We Simon Brooks Are the key Siam weed biological conduct investigations into the use Tropical Weeds Research Centre characteristics of age to maturity and of a splatter gun herbicide applicator. Tel: (07) 4761 5708 seed longevity the same in seasonally This equipment can be carried in Email: [email protected] drier, warmer areas as they are along the a backpack and relies on a higher Other staff in 2010–11 wet tropics coast? concentration of herbicide in a low- volume application. Shane Campbell, Wayne Vogler, Kirsty • Siam weed seeds sourced from Gough, Stephen Setter, Katie Patane and infestations in the wet and dry tropics How long will seeds remain viable when Sharon Rossow are germinated in the quarantine immersed in creek water and will this laboratories at the Centre for Wet survival influence search buffers? Is sea Objectives Tropics Agriculture (CWTA) and water a barrier to the dispersal of viable TWRC and seedlings are planted in Siam weed and limnocharis seeds? • Investigate key ecological attributes pots at monthly intervals for one year. influencing the eradication of species • All national weed eradication target We collect data on growth rates and targeted by national cost-share species occur along creek lines. In a flowering behaviour and the plants eradication programs. seed immersion trial in the ecology are destructively harvested as they laboratory at CWTA, we compare • Refine eradication methods by using mature. ecological information. the seed viabilities of L. flava, • We bury fresh Siam weed seed from C. hirta, M. calvescens and two seed • Investigate alternative control methods an infestation near Townsville in collections of Siam weed immersed for remote Siam weed infestations in permeable mesh packets to investigate for 2–126 days in creek water, sea the seasonally dry tropics. the effects of burial duration, grass water and a 50/50 mix. Results will Rationale cover, soil type and burial depth provide baseline data for models of in a dry tropics environment at aquatic dispersal. We also conduct Siam weed (Chromolaena odorata) is a TWRC. Packets are retrieved every laboratory investigations into seed Class 1 declared weed in Queensland 6–12 months until no viable seed buoyancy and seed germination under and has been the target of a national is recovered. Previous research has increasing levels of salinity. cost-share eradication program since its shown that viable seed is exhausted How persistent are L. flava, M. calvescens discovery in the wet tropics region of 7 years after burial in a wet tropics and C. hirta soil seed banks? northern Queensland in 1994. It has also environment. been found in the upper Herbert River • Soil cores are collected regularly catchment (1997) and in the upper Ross Can larger Siam weed plants be controlled from an area with a high density of River and Black River catchments west by repeated burning and how do repeat M. calvescens (prior to control) at of Townsville (2003). fires influence the soil seed bank? the El Arish infestation. Seed bank The National Four Tropical Weeds • We maintain monitoring plots studies for L. flavaare continuing at Eradication Program commenced in within a large Townsville Siam an infestation in a constantly wet, 2003, targeting four genera of Class 1 weed infestation and subject them spring-fed, lowland tropical stream weeds (Clidemia hirta, Limnocharis to repeated controlled burns. near Feluga. This data will guide the flava, Miconia calvescens, M. nervosa, Pre-burn data collected includes duration of control and monitoring M. racemosa and Mikania micrantha), plant size, fuel loads, soil seed banks activities, as there is no other which are located primarily on the wet and soil moisture levels. Post-burn information available on limnocharis tropics coast of Queensland. assessments include fire damage and seed persistence. All soil samples These eradication programs will only mortality of Siam weed, sizes of soil are sieved to remove, count and be successful if field crews can locate seed banks and seedling recruitment. germinate seed. all individuals of each species, apply Can Siam weed be effectively treated • We also bury C. hirta and effective control measures, prevent new with a low-volume, high-concentration M. calvescens seed in permeable seed production and monitor infestations herbicide application through a splatter mesh packets to investigate the until the seed bank is exhausted. To (gas) gun? effects of burial duration and depth address queries from the eradication in a wet tropics environment at programs, our research concentrates on CWTA. Packets are retrieved every key biological parameters (soil seed bank 6–24 months until no viable seed is persistence, age to maturity, flowering recovered. behaviour, seed production and dispersal vectors and barriers) and on effective control methods for each species.

Part 2 Landscape protection and restoration 39 Progress in 2010–11 Siam weed—fire response Limnocharis—seed bank persistence Siam weed—age to maturity For monitoring results following the first controlled burn in October 2008, Monitoring between 2003 and 2009 has At TWRC all of the Siam weed plants see Technical highlights 2008–09 shown no notable decline in viability raised from Townsville seed in late and Technical highlights 2009–10. or seed density of limnocharis in the October, November and December 2009 Controlled burns were not conducted soil seed bank at a site near Feluga flowered in May 2010. Also, 5 out of in the following two years due to (Table 18.1), even though seed input 6 plants raised from seed in late January dangerously dry conditions in 2009 and has not been recorded since 2004 and 2010 flowered in late May and were wet conditions in 2010. In June 2011, the number of plants emerging at the destructively harvested. All remaining we completed pre-burn assessments infestation has declined over this period. plants sown monthly until October and conducted a second controlled burn However, sampling in 2010 has detected 2010 flowered in May 2011. Some plants shortly after. a marked reduction in seed density from three additional treatments sown (Table 18.1). fortnightly from mid-January 2011 also Siam weed—splatter gun trials commenced flowering in May 2011. For results from the first trial in Additional research In the wet tropics trial, 6 out of March 2009, see Technical highlights In 2011, Siam weed leaf samples were 18 plants raised from seed between early 2009–10. A second trial in April 2010 collected from Townsville and wet November 2009 and mid-January 2010 demonstrated that lower rates of two tropics infestations and from a smaller, flowered in May 2010. All remaining fluroxypyr-based herbicides were also phenotypically different Siam weed plants that were sown monthly until effective. In May 2010, we conducted infestation near Tully. Collaborator November 2010 flowered in early a third trial to verify the effectiveness Dr Jane Oakey used the leaf samples May 2011. of lower rates of fluroxypyr compared to develop a specific test for low to higher rates of metsulfuron-methyl. concentrations of C. odorata DNA of Both trials show that under ideal (pot) Results are pending. both phenotypes. This investigation conditions, Siam weed can be raised may lead to the use of water sampling to from seed between November and Seed immersion trials determine whether Siam weed is locally January and flower by June. It is not yet present or absent. known how often plants would flower so Results of the seed immersion trial quickly under field conditions. However, demonstrated that 126 days of immersion A small trial was established in both trials highlight the importance of in any of the three water types tested collaboration with CSIRO Ecosystem effectively surveying and controlling is not a barrier to the dispersal of Sciences to measure the monthly infestations between early February and viable seed for all four species included emergence, survival and growth of flowering in May/June. Flowering plants in this trial. However, further trials M. racemosa plants in the field. More varied in size depending on their age, have shown that Siam weed seed will than 80% of seed obtained from a so this trial illustrates that the time to not survive when it is maintained in M. racemosa plant in the field has been flower bud initiation is driven by day solutions containing more that 250 mmol found to be viable and plants from length. of sodium chloride and M. calvescens these are currently raised to enable seedlings will not establish in solutions observations of age to maturity under Siam weed—seed longevity containing more than 100 mmol of controlled conditions. sodium chloride. So although seed can Seeds were buried in December 2009. We be dispersed by sea or brackish water, recorded a mean seed viability of 72% at Funding in 2010–11 highly saline environments are not the beginning of the trial. Considerable suitable habitats for plant establishment. • Queensland Government germination was noted in surface- • National Siam Weed Eradication situated bags during the wet season Miconia calvescens—seed bank Program and Four Tropical Weeds and seed viability averaged 23% after persistence Eradication Program, through 6 months and 12.5% after 12 months national cost-share arrangements across all burial depths. Viability after Sampling of the soil seed bank at ($60 000) 12 months of surface seed, seed buried at El Arish shows that small numbers of 3 cm and seed buried at 13 cm was 4.1%, seedlings continue to emerge and there 15.6% and 17.9% respectively. We are has been no decline in the density of currently assessing seed retrieved after seed extracted from the soil. 18 months. We also established a buried seed trial in October 2010. Prior to burial, over 80% of seed was viable. We are currently assessing seed retrieved after 6 months.

40 Technical highlights: research projects 2010–11 Table 18.1 Number, distribution and viability of L. flava seeds extracted from 40 mud samples collected between 2003 and 2010 at an infestation near Feluga, northern Queensland

Year 2003 2005 2006 2007 2008 2009 2010 Number of seeds sieved from 40 samples 623 358 1252 489 530 323 8 Samples with seeds (%) 70.7 77.5 82 55.5 67.5 47.5 2.5 Average number of seeds per sample 15.2 9.0 31.3 12.2 13.3 8.1 0.4 Seed viability (%) — 64.7 54.4 59.5 80.7 57.9 100

Collaborators More information • Biosecurity Queensland officers based For further information on this at South Johnstone and Townsville— research project, visit the invasive provided assistance with locating and plant and animal science pages on the accessing trial areas Biosecurity Queensland website at www.biosecurity.qld.gov.au • DERM staff—coordinated the controlled Siam weed burn • Dr Jane Oakey, Molecular Biologist (Biosecurity Queensland) • CSIRO Ecosystem Sciences, Atherton • School of Land, Crop and Food Sciences, The University of Queensland

Part 2 Landscape protection and restoration 41 19. Class 1 weed control packages

Project dates Control efforts are also hindered by a Although their flowering was prolific, lack of chemical registrations. Of the water mimosa (Neptunia oleracea) July 2008 – June 2013 53 naturalised Class 1 plant species, failed to produce seed at the Alan Project leader 9 have chemical recommendations, Fletcher Research Station, Ecosciences 21 are captured under broader Precinct and TWRC, delaying planned Joseph Vitelli categories (e.g. Acacia spp., cacti or seed studies. To resolve Neptunia Ecosciences Precinct the Environmental Permit PER11463) identification problems in Queensland, Tel: (07) 3255 4473 and 23 have no chemicals registered. N. oleracea and N. plena DNA samples Email: [email protected] A minor use herbicide permit covering were sought from the New York all Queensland Class 1 declared plant Herbarium and the Kew Royal Botanic Other staff in 2010–11 species would greatly aid eradication Gardens Herbarium. Molecular probes Barbara Madigan and Annerose efforts and enhance Queensland’s ability for these two species have now been Chamberlain to respond to new incursions. developed by Dr Jane Oakey.

Objectives Methods Eight additional Cecropia sites have been found in northern Queensland. Develop reliable and effective control Ecological studies Growth scars indicate that some plants options that can be integrated into are more than 20 years old. At a Clifton We collect basic ecological information eradication programs for Queensland Beach location, three separate cohorts (e.g. flowering period, seed production, Class 1 declared weeds. This includes: have established, with plant height age to reproductive maturity and seed ranging from 1.3 to 14 m. Morphological • seeking a minor use herbicide permit bank persistence) on prioritised species traits indicate at least three Cecropia that covers all Queensland Class 1 from established infestations. All seeds species have established in northern declared plant species are collected and removed from sites. Queensland. We have collected leaf At the conclusion of the study, all plants • collecting basic ecological data samples from most sites and have stored are killed. (e.g. time to reproductive maturity them for future DNA analysis. and soil seed bank dynamics) on priority Class 1 weeds Control studies Accelerated seed ageing trials have been initiated for Senegal tea, Siam weed, We develop chemical and non-chemical • implementing an accelerated ageing Acacia glauca, Koster’s curse, miconia control options that can be integrated test to determine potential seed and badhara bush. Germination/viability into eradication programs and provide longevity on targeted Class 1 weeds testing is ongoing. (where more than 1200 mature seeds data to assist minor use applications. can be sourced). Progress in 2010–11 Control studies Rationale An alligator weed (Alternanthera Ecological studies philoxeroides) herbicide trial at Bullock Expanding the current Queensland A seed library of Class 1 weeds Head Creek using glyphosate at two Class 1 plant declaration list to species naturalised in Queensland to date and three times the recommeneded rate level yielded a list of 8978 taxa, contains seeds from Mexican feather appears to have been effective. The including hybrids, cultivars and grass (Nassella tenuissima), Mexican dense stand of alligator weed along 1 km synonyms. Restricting the list to bean tree (Cecropia peltata), trumpet of the creek has been reduced to isolated those Class 1 species already present 2 tree (Cecropia palmata), badhara bush plants totalling less than 2 m . in Australia narrowed the total to (Gmelina elliptica), mimosa (Mimosa 156, of which 53 species are currently A foliar spray screening trial involving pigra), Koster’s curse (Clidemia hirta), naturalised in Queensland. For the five herbicides was initiated on Senegal miconia (Miconia calvescens), Senegal majority of these species there is tea growing at a Strathpine trial site in tea (Gymnocoronis spilanthoides), Siam anecdotal, limited or no ecological April 2011. weed (Chromolaena odorata) and Acacia information (e.g. age to reproductive glauca. Class 1 minor use permit maturity and seed bank persistence) available. Such data is essential for Field monitoring studies showed that the A draft minor use permit capturing all enhancing the effectiveness and reproductive period for Mexican feather Queensland Class 1 declared weeds is efficiency of eradication efforts. grass is from September to March. We now complete. The final version will be Furthermore, knowledge of seed bank have transplanted 38 seedlings of known submitted to APVMA by May 2012. persistence provides insights into how age from the field and repotted them at To assist Class 1 eradication efforts, long programs need to be maintained the Ecosciences Precinct to determine we obtained new minor use permits for and the resources needed to achieve time to reproductive maturity. To date, control of gingers (Hedychium spp.), eradication. 16% of plants have flowered at an age PER12436, and tussock grasses of 19–23 months between August and (Nassella spp.), PER12202. (For details, March. Studies are ongoing. see project report 34 on page 76.)

42 Technical highlights: research projects 2010–11 Photo 19.1 (a) Inflorescence, (b) pod, (c) adult tree, (d) leaves, (e) trunk and (f) seed of Acacia pringlei, a newly detected non-indigenous acacia species; all non-indigenous Acacia species are Class 1 declared weeds in Queensland

Funding in 2010–11 Queensland Government More information For further information on this Collaborators research project, visit the invasive • Dr Jane Oakey, Molecular Biologist plant and animal science pages on the (Biosecurity Queensland) Biosecurity Queensland website at www.biosecurity.qld.gov.au • Biosecurity Queensland field staff • Brisbane City Council • Capricorn Pest Management Group • Logan City Council • Seqwater

Part 2 Landscape protection and restoration 43 20. Mimosa pigra research

Project dates Methods Africa, North, Central and South America, and the Asia–Pacific region. July 2008 – June 2013 The study site at Peter Faust Dam is located on the peninsula known as The level of homogeneity found within Project leader Point 10, extending from the 65% water the Australian samples was high, Joseph Vitelli storage capacity level to the middle of suggesting that a low level of genetic Ecosciences Precinct the creek bed. This area includes closed- diversity exists for this species within Tel: (07) 3255 4473 canopy M. pigra infestations (known Australia. This implies that there might Email: [email protected] as core areas) and individual M. pigra have been only a single introduction of plants scattered across the peninsula. M. pigra into Australia. Results from the Other staff in 2010–11 neighbour-joining tree analysis indicate Seedling emergence and seed bank that the Australian population cluster Barbara Madigan We record seedling counts annually in may be related to the cluster containing Objectives a 5 m grid pattern across the peninsula specimens from Brazil and one region in and take soil cores annually from Mexico (Guerrero, San Marcos). Further • Study seedling emergence and different areas for seed bank studies. We specimens will need to be examined seed bank dynamics of Mimosa also test the viability of recovered seeds. before any credence is given to this pigra growing at Peter Faust Dam, relationship. The work is ongoing. Proserpine, to assist in the eradication Molecular studies of this species. Funding in 2010–11 Next-generation (whole-genome) • Investigate fire and chemical options sequencing of M. pigra DNA is Queensland Government for M. pigra control. undertaken by the Australian Genome Collaborators • Determine the origins of M. pigra Research Facility using the 454 infestations in Queensland, Western sequencing system. Then Dr Jane Oakey • Dr Jane Oakey, Molecular Biologist Australia and the Northern Territory uses up to 150 microsatellite loci for (Biosecurity Queensland) M. pigra to determine alleles from through next-generation (whole- • Kay Bailey, National WONS Coordinator M. pigra infestations in Queensland and genome) sequencing. for M. pigra and athel pine, and Western Australia and from 20 reference Christopher Collins, Bert Lukitsch, Ian populations in the Northern Territory. Rationale Cowie, Ben Stuckey, Em Pedler and Ben Mimosa pigra is a WONS and a In addition, we analyse populations from Matthias (NRETAS, Northern Teritory) Class 1 declared weed in Queensland. Africa (Botswana and Kenya), North, • Tim Heard and Gio Fichera (CSIRO Originating from Central America, Central and South America (Brazil, Ecosystem Sciences) M. pigra poses a major threat to the Columbia, Costa Rica, Cuba, Ecuador, integrity of northern Australia’s El Salvador, Guyana, Nicaragua, Venezuela • Chris Hawkins and Tracey Vinnicombe wetlands, reducing biodiversity and and Mexico) and the Asia–Pacific region (Department of Agriculture and Food, affecting primary production. In (Indonesia, Malaysia, Papua New Guinea, Western Australia) the Northern Territory it has formed Thailand and Vietnam). The project • Syd Laker (Adjumarllarl Rangers) impenetrable, nearly mono-specific aims to process up to 200 samples. This • Willy Rioli, Vivian, Kim, D. Tipakalippa, thickets over 800 km2. In February information will then be used to evaluate Colin and Nicholas Hunter (Tiwi Land 2001, the first infestation of M. pigra and compare allele frequencies between Council, Pirlingimpi Community) in Australia outside the Northern populations and determine the most Territory was found at Peter Faust likely source of each of the Australian More information Dam, near Proserpine in central coastal populations. Queensland. A stakeholder group was Key publication formed to eradicate the infestation. Progress in 2010–11 Vitelli, JS, Madigan, BA & Worsley, KJ One of Biosecurity Queensland’s 2006, ‘Mimosa pigra in Queensland’, in contributions is to provide research on Seedling emergence and seed bank C Preston, JH Watts and ND Crossman the biology and control of M. pigra to aid Soil cores were not extracted during (eds), Proceedings of the 15th Australian in the eradication effort. This includes 2010–11, as the core area remains Weeds Conference, Weed Management advising on the timing of site revisits to inundated as a result of the heavy rains Society of South Australia, Adelaide, ensure plants are detected and controlled in early 2010. pp. 251–4. prior to setting seed, and predicting how long the eradication effort needs Molecular studies For further information on this research to continue. project and access to key publications, We isolated 215 polymorphic visit the invasive plant and animal science microsatellite markers from an enriched pages on the Biosecurity Queensland library of M. pigra genome DNA and website at www.biosecurity.qld.gov.au used 76 of these microsatellite loci to screen 234 samples from Australia,

44 Technical highlights: research projects 2010–11 21. Ecology and control of wet tropics weeds

Project dates Pond apple mechanical control Bog moss January 1999 – June 2013 Mechanical control of pond apple may We undertake field trials in a small be a viable option in some areas during drain infested with bog moss to provide Project leader drier periods. We test two different control options for this emerging Melissa Setter machines—the Positrack™ and the aquatic weed. Treatments include the –1 Tropical Weeds Research Centre Tracksaw™—for their kill rate, amount use of herbicides (UniMaz 250™ 4 L ha –1 Tel: (07) 4064 1149 of follow-up control required, cost- with Bonus™ wetter 1.5 L ha ) as well Email: [email protected] effectiveness and selectivity (effect on as shadecloth (92% light reduction) native vegetation). Trials are performed and black plastic (for more sensitive Other staff in 2010–11 in pond apple infestations of similar size environments). and density at the Daintree (Positrack™) Stephen Setter, Katie Patane, Wayne We also conduct a herbicide screening trial and in Innisfail (Tracksaw™). There are Vogler, Laura Roden and Barbara in the glasshouse, in which 10 herbicides notable differences between the two Madigan (triclopyr, endothal, glyphosate, imazapyr, machines: the Positrack™ creates mulch metsulfuron, dichlobenil, flumioxazin, Objective from the destroyed plants, while the bispyribac, carfentrazone and diquat Tracksaw™ has herbicide application + guar gum) are added at two rates to Increase our understanding of the integrated into its control method and ecology and control options of key submerged bog moss growing in 2 L jars. thus requires only a single operator. (For The most effective herbicides are then wet tropics weeds to improve their the Positrack™ experiment, we applied management. tested in a rates trial in jars and in a large herbicide by hand.) container trial with emergent bog moss. Rationale Navua sedge Progress in 2010–11 Weeds pose a major threat to the high We conduct a foliar herbicide screening economic, environmental and social trial near Millaa Millaa on the Atherton Pond apple mechanical control values of land in the wet tropics. Many Tableland and we use the results to We have established transects and wet tropics weeds are relatively recent determine the next phase of herbicide applied the mechanical control arrivals and have not reached the trials. The aim is to develop effective treatments; monitoring and analysis are full extent of their range and impact. selective herbicide options for navua ongoing (Table 21.1). Very wet conditions Much of the basic ecological knowledge sedge control in sown tropical pastures prevented access, so we could not control required to develop comprehensive and non-agricultural areas. regrowth as planned. Fruiting has been long-term control strategies for wet observed on resuckering trees 18 months tropics weeds is unavailable. This after treatment. project conducts field, shadehouse and laboratory experiments on a number of priority weed species. Research findings will enable land managers to more effectively limit weed impacts on natural ecosystems, primary industries and tourism.

Methods Field, shadehouse and laboratory experiments are currently underway on a number of weed species, including pond apple (Annona glabra), navua sedge (Cyperus aromaticus) and bog moss (Mayaca fluviatilis).

Photo 21.1 Stephen Setter applying herbicides to a navua sedge plot as part of initial screening trials

Part 2 Landscape protection and restoration 45 Table 21.1 Area treated, cost and efficacy of the PositrackTM and TracksawTM machines in controlling pond apple and regrowth 12 months after treatment

PositrackTM TracksawTM Area treated in 2 days (ha) 1.5 0.75 Initial pond apple density (plants ha–1) 11 000 3 300 Mortality—mechanical control only (%) 83 n/a Mortality—mechanical + herbicide control (%) 95 90 Treatment cost ($ ha–1) 2 300 2 730 Regrowth 12 months after treatment (seedlings ha–1) 2 600 18 500

Table 21.2 Herbicide efficacy on navua sedge in the foliar spray screening trial at Millaa Millaa 2010

Live stems (% of baseline) Product application 42 days after 84 days after 140 days after Treatmenta Active ingredient rate treatment treatment treatment Control — — 140 113 200 Sempra™ halosulfuron 130 g ha–1 60 103 236 Sempra™ + halosulfuron + glyphosate 130 g ha–1 + 3 L ha–1 37 37 81 Roundup 360™ Sempra™ + halosulfuron + 2,4-D amine 130 g ha–1 + 3.2 L ha–1 56 143 232 Amicide 625™ Sempra™+ halosulfuron + 2,4-D amine 130 g ha–1 + 3.2 L ha–1 + 37 81 167 Amicide 625™ + + dicamba 1.44 L ha–1 Kamba 500™ Sempra™ + halosulfuron + 2,4-D amine 130 g ha–1 + 3.2 L ha–1 + 50 115 223 Amicide 625™ + + ammonium sulphate 28.8 L ha–1 Liase™ Sempra™ + halosulfuron + MCPA amine 130 g ha–1 + 2.67 L ha–1 64 154 314 Agritone™ NUL2452b NUL2452 130 g ha–1 45 62 178 NUL2452 + LVEc NUL2452 + LVE MCPA 130 g ha–1 + 1.75 L ha–1 47 78 208 Agritone™ NUL2452 + NUL2452 + glyphosate 130 g ha–1 + 3 L ha–1 47 25 79 Roundup 360™ Amicide 625™ + 2,4-D amine + dicamba 3.2 L ha–1 + 1.44 L ha–1 82 118 225 Kamba 500™ Amicide 625™ 2,4-D amine 3.2 L ha–1 84 117 181 Arsenal Express™ imazapyr + glyphosate 7 L ha–1 40 18 48 Flame™ imazapic 1 L ha–1 68 121 268 Flame™ + imazapic + glyphosate 1 L ha–1 + 3 L ha–1 46 23 151 Roundup 360™ Midas™ MCPA + imazapic + 4 L ha–1 82 256 305 imazapyr Arsenal 250™ imazapyr 4 L ha–1 46 63 154 a Pulse adjuvant has been added to all herbicides at 2 mL L–1. b NUL2452 is a Nufarm experimental product. c LVE = low-volatility ester.

46 Technical highlights: research projects 2010–11 Navua sedge More information The foliar herbicide screening trial has Key publications been completed. No herbicide treatment adequately controlled navua sedge. Westcott, DA, Setter, MJ, Bradford, MG, All herbicides reduced the number McKeown, A & Setter, S 2008, of navua sedge live stems at 42 days ‘Cassowary dispersal of the invasive after treatment. These numbers had pond apple in a tropical rainforest: the increased at 84 days after treatment contribution of subordinate dispersal in all treatments except those modes in invasion’, Diversity and containing glyphosate, in which live Distributions 14(2): 432–9. stems continued to decline. However, Mason, LB, Setter, MJ, Setter, SD, at 140 days after treament, live stem Hardy, T & Graham, MF 2008, ‘Ocean numbers increased in all treatments, dispersal modelling for propagules with those not containing glyphosate of pond apple (Annona glabra L.)’, in showing consistently more than 150% of RD van Klinken, VA Osten, FD Panetta the baseline value (Table 21.2). & JC Scanlan (eds), Proceedings of the Although treatments containing 16th Australian Weeds Conference, glyphosate were the most effective, Queensland Weeds Society, Brisbane, they are not selective and therefore not Queensland, pp. 519–21. suitable for use in grass-based pasture Setter, SD, Setter, MJ, Graham, MF systems and other non-agricultural & Vitelli, JS 2008, Buoyancy and situations where selective herbicides germination of pond apple (Annona are required. Future research will glabra L.) propagules in fresh and salt focus on Sempra™ and Sempra-based water’, in RD van Klinken, VA Osten, herbicide mixes, including split herbicide FD Panetta & JC Scanlan (eds), application in combination with Proceedings of the 16th Australian Weeds pasture management practices. Other Conference, Queensland Weeds Society, herbicides will be screened for efficacy Brisbane, Queensland, pp. 140–2. as necessary. For further information on this Bog moss research project and access to key publications, visit the invasive plant In the field trials, estimated biomass and animal science pages on the reduction 120 days after treatment Biosecurity Queensland website at for the herbicide, black plastic and www.biosecurity.qld.gov.au shadecloth treatments were 100%, 100% and 95% respectively. For the herbicide screening trial, we made final assessments at 120 days after treatment. We will now test the six most effective herbicides (triclopyr, endothal, diquat + guar gum, metsulfuron, fumioxazin and carfentrazone) at four rates each in a rates trial.

Funding in 2010–11 • Land Protection Fund ($62 000) • Queensland Government

Collaborators • Cairns Regional Council • Cassowary Coast Regional Council • Far North Queensland Regional Organisation of Councils

Part 2 Landscape protection and restoration 47 22. Population viability analysis models for better management of lantana (Lantana camara)

Project dates Methods Progress in 2010–11 July 2008 – June 2012 We set up and census permanent plots of The third and final census of population 50 m × 50 m at each of four infested sites dynamics across the four field Project leader in the Yarraman–Blackbutt area west of sites—consisting of more than 2000 Dr Olusegun Osunkoya Brisbane (Photo 22.1) to determine the permanently tagged lantana plants—was Ecosciences Precinct parameters of lantana’s vital rates (seed completed. We are now using census Tel: (07) 3255 4474 germination and dormancy, seedling, data to explore spatio-temporal variation Email: [email protected] juvenile and adult growth and survival, in stand dynamics, spatial patterns and and adult fecundity) for projection of its population growth of the weed. population growth. We also use spatial Other staff in 2010–11 Populations at most sites exhibited plant analytical techniques to assess first- Christine Perrett and Cameron Clark height and crown size distributions that order and second-order spatial patterns are L‑shaped, J‑shaped or symmetrical of established individuals and recruits. Objectives and thus unimodal. Bimodality was The four field sites (hoop pine plantation, observed at the farm site only; this • Use size-structured population natural forests subject to either periodic suggests that two distinctive cohorts matrices and mathematical models burning or grazing regimes, and cattle exist (Figure 22.1). Relative growth rate to examine vital rates of growth, property) contain, in increasing order, low of lantana plants at the farm site was reproduction and survival of lantana to moderate infestations of lantana, but not influenced by size. This contrasts under various landscape scenarios to differ in soil properties, rainfall intensity, significantly with trends observed in the project its population growth into the land use type and weed control practices. other three sites, where one-sided (size future. asymmetric) growth and competition • Identify, from a suite of demographic were apparent. parameters and with the aid of computer simulations and model Farm Forest (grazing) Forest (burning) Hoop pine plantation

predictions, the main driver/s 2008 2008 50 2008 2008 30 120 60 of population growth that could 25 100 40 50 be manipulated for management 20 80 30 40 15 60 30 20 purposes. 10 40 20 10 5 20 10 0 • Analyse spatio-temporal changes in 0 0 0 0 100 200 300 0 100 200 300 0 100 200 300 0 100 200 300 400 stand dynamics of lantana to explore 2009 2009 2009 2009 50 60 30 120 better management of the weed. 50 25 100 40 80 40 20 30 30 15 60 Rationale 20 20 10 40 Frequency 10 10 5 20 Lantana (Lantana camara) is a WONS 0 0 0 0 and a Class 3 declared weed in 0 100 200 300 0 100 200 300 0 100 200 300 0 100 200 300 400 Queensland. Despite millions of dollars 2010 2010 2010 30 120 50 spent on control, research and extension 25 100 40 20 80 work, there is a dearth of quantitative 30 15 60 20 data encompassing the entire life cycle 10 40 5 20 10 of the weed. To date, no attempt has been 0 0 0 0 100 200 300 made to carry out population viability 0 100 200 300 0 100 200 300 analyses on the species, despite the Plant height (cm) widely held view that such analyses, when done in concert with sensitivity Figure 22.1 Changes in lantana plant height distribution in each of the four populations surveyed over three years: note that the y-axis (frequency) scale varies between sites analyses and numerical simulations, could help greatly in finetuning management strategies for control of invasive organisms. This project aims to fill this apparent gap in our understanding of the invasion biology of lantana.

48 Technical highlights: research projects 2010–11 Spatial pattern analyses indicated 5.0 that established and newly recruited individuals are aggregated and that the 4.5 degree of aggregation decreases with 4.0 plant size. However, after fire, seedling and juvenile recruitment assume negative 3.5 farm association (spatial displacement/ 3.0 decoupling) in relation to established forest (grazing) individuals, perhaps in response to 2.5 increased lantana litter (fuel load) and thus 2.0 forest (burning) more intense temperature build-up around (log) abundance Fruit 1.5 existing parent plants during burning. hoop pine plantation This finding implies that in a landscape 1.0 where burning is used as a management tool, any follow-up with herbicide that 0.5 focuses on reducing recruitment can 0.0 simply concentrate on spaces between 1.5 1.8 2.1 2.4 established individuals, thereby reducing Plant height (log cm) both herbicide and labour cost. Figure 22.2 Lantana fruit production in 2008 as a function of plant height in the four populations surveyed Density and plant size had appreciable effects on the weed’s reproductive farm capacity and growth, but surprisingly forest (grazing) (a) 100 not on survival. The trends in fruit forest (burning) production as a function of plant 80 hoop pine height were statistically significant at plantation p < 0.05. The intercept but not the slope 60 differed significantly (p < 0.05) between 40 populations (Figure 22.2). 20

On average, lantana populations exhibit (%) flowering plants Lantana 0 reproductive activity at least twice per Feb. Apr. June Aug. Oct. Dec. Feb. Apr. June Aug. Oct. Dec. Feb. Apr year, and even more in a good season, 2009 2010 2011 Date when resources (especially moisture) are adequate (Figure 22.3). The lantana (b) 100 population at the hoop pine plantation could only be followed in the first 80 10 months of the survey due to closing 60 of the canopy and the impenetrable thicket that then developed. 40

20 Lantana plants fruiting (%) fruiting plants Lantana

0 Feb. Apr. June Aug. Oct. Dec. Feb. Apr. June Aug. Oct. Dec. Feb. Apr 2009 2010 2011 Date

Figure 22.3 Pattern of (a) flowering and (b) fruiting of lantana over two years in the four populations surveyed

Part 2 Landscape protection and restoration 49 Periodic but slow/moderate burning of (a) 2008 Significance 5000 2009 Site p < 0.02 a lantana infestation as a management 2010 ) Year p < 0.02 tool appears not to tilt the population –2 standard error Site × year p < 0.001 to negative growth, although it can 4000 reduce soil-stored seed viability and abundance (which can be as high as –2 3000 1000 seeds m ). Approximately 25% of fresh lantana seeds remain viable even after three years of burial, confirming 2000 that seed persistence of the weed can be long (Figure 22.4). 1000 Environmental variability is the norm

Soil-stored seed abundance (seeds m (seeds seed abundance Soil-stored rather than the exception. Over three 0 years, our surveys across varying Farm Forest (grazing) Forest (burning) Hoop pine plantation landscape and land-use types have Site enabled us to capture such variability.

2008 Significance The next task is to build robust (b) 2009 Site p < 0.05 population growth models and then 12 2010 Year Not significant combine the demographic information standard error Site × year p < 0.05 collected with economic data 10 (e.g. control cost per plant or per hectare) and environmental data (e.g. long-term 8 rainfall trend). This will lead to better informed decisions on the feasibility of 6 local control/eradication of the weed.

4 Funding in 2010–11

2 • Queensland Government Soil-stored seed germination (%) seed germination Soil-stored • Land Protection Fund ($37 000) 0 Farm Forest (grazing) Forest (burning) Hoop pine plantation Site Collaborators • S. Raghu (CSIRO Ecosystem Sciences, fresh seeds Significance Brisbane) (c) 1-year-old seeds Site p < 0.02 50 2-year-old seeds Year p < 0.05 • Joe Scalan (Biosecurity Queensland, standard error Site × year p < 0.05 45 Toowoomba) 40 More information 35 Key publication 30 Osunkoya, OO, Perrett, C & Fernando, C 25 2010, ‘Population viability analysis 20 models for Lantana camara L.

Seed germination (%) Seed germination 15 (Verbenaceae): a weed of national significance’, in SM Zydenbos (ed.), 10 Proceedings of the 17th Australasian 5 Weeds Conference, New Zealand Plant 0 Protection Society, Christchurch, New Farm Forest (grazing) Forest (burning) Hoop pine plantation Zealand, pp. 99–102. Site For further information on this research Figure 22.4 Seed trait dynamics of lantana in each of the four populations surveyed: (a) soil- project and access to key publications, stored seed abundance, (b) soil-stored seed germination and (c) seed germination in relation to time since soil burial visit the invasive plant and animal science pages on the Biosecurity Queensland website at www.biosecurity.qld.gov.au

50 Technical highlights: research projects 2010–11 23. Impacts of environmental weeds on soil processes

Project dates • leaves of cat’s claw creeper and In soils supporting cat’s claw creeper, co-occurring native vines (Parsonsia the following were significantly January 2010 – June 2012 straminea and Smilax australis) elevated: moisture, pH, calcium, organic Project leader and introduced but low-impact carbon (OC), total carbon (TC), total vines (corky passionfruit, Passiflora nitrogen (TN) [but not exchangeable Dr Olusegun Osunkoya suberosa) to explore differences in nitrogen in the form of ammonium or Ecosciences Precinct leaf chemistry between invasive and nitrate], magnesium, sulfur, chloride Tel: (07) 3255 4474 non-invasive species. and electrical conductivity (EC). Email: [email protected] Collected data is then subjected to Exchangeable sodium and iron (which can be toxic to plant growth if present Other staff in 2010–11 univariate as well as multivariate ordination techniques. in excessive levels) were present at Christine Perrett and Cameron Clark lower levels (Figure 23.1). These results Progress in 2010–11 indicate that cat’s claw creeper—like Objective lantana (see Technical highlights Document the impacts of environmental Invasion effect on soil ant 2009–10 and Osunkoya & Perrett 2011)— weeds on soil processes. assemblage can improve soil fertility and influence nutrient cycling, making the substratum The epigaeic and subterranean strata ideal for its own growth. This might Rationale supported markedly different ant explain the ability of the weed to out- The presence of aggressive exotic assemblages, and ant communities also compete other species, especially plants poses a serious threat to natural differed between riparian and non- native ones. ecosystems. Lantana (Lantana camara) riparian habitats. However, cat’s claw creeper invasion had a limited impact. and various weedy vines—such as cat’s Leaf chemistry of cat’s claw There was a tendency for ant abundance claw creeper (Macfadyena unguis-cati) creeper v. non-invasive species and Madeira vine (Anredera cordifolia)— and species richness to be lower in are exotic weeds of great significance in infested patches, and overall species As for the soil data, leaf chemistry of Queensland and nationally. Currently, composition was different between cat’s claw creeper and non-invasive there is little quantitative information on infested and uninfested patches, but species varied more across survey changes in below-ground (soil) ecosystem these differences were relatively small, sites than between invasion status properties mediated by the presence of and did not occur consistently (Table 23.2). Nonetheless, following these invasive species, including changes across sites. removal of the site effect, about 75% of nutrients tested in cat’s claw creeper in physicochemical properties, insect Our study has shown that ant were significantly different from diversity and microbial activity. communities are surprisingly resilient those in low-impact introduced vines. to invasion by cat’s claw creeper, and Cat’s claw creeper leaves possess more Methods serves as a warning against making nitrogen, phosphorus, potassium, At eight cat’s claw creeper field sites of assumptions about invasion impacts aluminium, boron, cadmium, cobalt, varying habitat and land-use type in the based on visual appearances. A paper copper, iron and molybdenum but Brisbane – Gold Coast region, we sample: detailing research findings has been significantly less calcium, magnesium published in Biological Invasions • surface active (epigaeic) and and manganese than the other vines. (Osunkoya et al. 2011). subterranean soil (hypogaeic) ants This trend of higher nutrient load in infested and uninfested areas Invasion effect on soil (especially nitrogen, phosphorus and potassium) for cat’s claw creeper is in spanning both riparian and non- physicochemical properties riparian habitats to investigate line with invasive plants’ preference patterns of ant species composition A significant site effect was more for disturbed habitats commonly and functional grouping in response frequently observed than an effect due associated with high pulses of nutrient to patch invasion status, landscape to invasion status. Nonetheless, after input. Calcium has been implicated in type and habitat stratum controlling for site differences, about formation of oxalate crystals in leaves 50% of the 23 soil traits examined for defence. The significantly lower level • soil beneath and away from differed significantly between infested of calcium in cat’s claw creeper leaves established cat’s claw creeper patches and non-infested soils (Table 23.1). despite a higher level of this nutrient in (4 samples per site) to document the the soil could be due to a reduced level likely effect of the weed on 23 soil of herbivory from natural enemies in physicochemical properties novel environments lowering the need to accumulate defence materials.

Part 2 Landscape protection and restoration 51 Table 23.1 Two-way analysis of variance (ANOVA) summary results of soil physicochemical properties in patches infested with and patches not infested with cat’s claw creeper across six sites in the Brisbane – Gold Coast region

Summary ANOVA Group meana (± SEb) Soil trait Site Invasion Site × invasion Infested Uninfested Air dry moisture (%) p ≤ 0.001 p ≤ 0.01 NSc 3.16 ± 0.28 2.54 ± 0.23 Soil macronutrients Organic carbon (%) p ≤ 0.001 p ≤ 0.01 NS 4.09 ± 0.34 3.48 ± 0.28 Total carbon (%) p ≤ 0.001 p ≤ 0.001 NS 4.74 ± 0.36 3.99 ± 0.28 Total nitrogen (%) p ≤ 0.001 p ≤ 0.001 NS 0.35 ± 0.03 0.28 ± 0.02 Exchangeable ions and micronutrients Nitrate nitrogen (mg kg–1) p ≤ 0.001 NS NS 17.71 ± 2.10 15.17 ± 1.30 Nitrate ammonia (mg kg–1) p ≤ 0.001 NS p ≤ 0.05 3.10 ± 0.50 3.40 ± 0.59 Phosphorus (mg kg–1) p ≤ 0.001 NS NS 29.38 ± 3.77 28.42 ± 5.06 Potassium (meq 100 g–1) p ≤ 0.001 NS NS 0.73 ± 0.07 0.62 ± 0.08 Calcium (meg 100 g–1) p ≤ 0.001 p ≤ 0.05 NS 15.69 ± 2.04 12.03 ± 1.79 Exchangeable sodium (meq 100 g–1) p ≤ 0.001 p ≤ 0.01 NS 0.05 ± 0.01 0.09 ± 0.02 Total sodium (meq 100 g–1) p ≤ 0.001 NS NS 0.11 ± 0.02 0.14 ± 0.02 Magnesium (meq 100 g–1) p ≤ 0.001 p ≤ 0.10 NS 5.55 ± 0.55 4.74 ± 0.59 Copper (mg kg–1) p ≤ 0.001 NS NS 1.29 ± 0.21 1.35 ± 0.23 Iron (mg kg–1) p ≤ 0.001 p ≤ 0.10 NS 87.10 ± 10.12 109.64 ± 10.12 Manganese (mg kg–1) p ≤ 0.001 NS NS 42.52 ± 7.42 41.46 ± 6.90 Zinc (mg kg–1) p ≤ 0.01 NS NS 4.27 ± 1.12 3.47 ± 1.15 Sulfur (mg kg–1) p ≤ 0.001 p ≤ 0.10 NS 12.54 ± 1.11 10.62 ± 1.37

Chemical characteristics pH p ≤ 0.001 p ≤ 0.01 NS 6.23 ± 0.14 5.90 ± 0.14 Electrical conductivity (dS m–1) p ≤ 0.001 p ≤ 0.001 NS 0.10 ± 0.01 0.07 ± 0.01 Chloride (mg kg–1) p ≤ 0.01 p ≤ 0.01 p ≤ 0.10 24.96 ± 0.78 18.60 ± 2.58 Soil texture Coarse sand (%) p ≤ 0.001 NS NS 23.83 ± 1.99 24.42 ± 2.41 Fine sand (%) p ≤ 0.001 NS NS 33.96 ± 2.14 35.58 ± 1.84 Silt (%) p ≤ 0.001 NS NS 19.50 ± 1.48 20.08 ± 1.75 Clay (%) p ≤ 0.001 NS NS 25.83 ± 1.68 23.83 ± 1.46 a Group means have been pooled across two habitats (riparian and non-riparian) and six survey sites. b SE = standard error. c NS = not significant.

52 Technical highlights: research projects 2010–11 4 Soil condition

Boonah Moogerah uninfested Boonah Moogerah infested 2 Ipswich

Ipswich

0 Nerang Nerang Oxley Oxley

–2 (S, Mn) Axis II [32%] (pH, P, Mg, K) II [32%] (pH, P, (S, Mn) Axis Carindale Carindale –4

–6 –6 –4 –2 0 2 4 Axis I [41%] (TN, OC, TC, Ca, Na, EC)

Figure 23.1 Ordination of soils underneath (closed symbols) and away from (open symbols) cat’s claw creeper patches across six sites of varying land use: arrows indicate direction and magnitude of soil traits driving each axis; percentages refer to data variation captured by each axis

Table 23.2 Two-way ANOVA summary results of leaf chemistry in cat’s claw creeper compared to co-occurring introduced non-invasive (P. suberosa) and native (P. straminea and S. australis) species across six sites in the Brisbane – Gold Coast region

Summary ANOVA Group meana (± SEb) Leaf trait Site Invasion Site × invasion Invasive Non-invasive Macronutrients (%) Total nitrogen p ≤ 0.001 p ≤ 0.001 NSc 2.88 ± 0.01 2.06 ± 0.09 Total phosphorus p ≤ 0.001 p ≤ 0.001 p ≤ 0.001 0.42 ± 0.03 0.26 ± 0.02 Potassium p ≤ 0.001 p ≤ 0.01 NS 2.45 ± 0.12 1.82 ± 0.11 Calcium NS p ≤ 0.01 p ≤ 0.05 1.65 ± 0.09 2.45 ± 0.24 Magnesium NS p ≤ 0.001 NS 0.26 ± 0.02 0.59 ± 0.07 Sodium p ≤ 0.001 NS p ≤ 0.01 0.04 ± 0.01 0.06 ± 0.01 Sulphur p ≤ 0.01 NS NS 0.19 ± 0.02 0.23 ± 0.02 Micronutrients (mg kg–1) Aluminium p ≤ 0.05 p ≤ 0.05 NS 82.38 ± 9.62 56.20 ± 9.28 Boron NS p ≤ 0.001 p ≤ 0.05 80.72 ± 4.55 40.41 ± 3.09 Cadmium p ≤ 0.001 p ≤ 0.01 NS 0.04 ± 0.01 0.02 ± 0.01 Cobalt p ≤ 0.001 p ≤ 0.001 p ≤ 0.001 0.26 ± 0.02 0.16 ± 0.02 Copper p ≤ 0.001 p ≤ 0.05 p ≤ 0.01 17.14 ± 1.62 13.47 ± 1.23 Iron p ≤ 0.001 p ≤ 0.001 NS 114.82 ± 11.38 75.86 ± 9.80 Manganese p ≤ 0.001 p ≤ 0.001 p ≤ 0.001 70.79 ± 80.5 186.21 ± 69.78 Molybdenum p ≤ 0.001 p ≤ 0.001 p ≤ 0.001 1.56 ± 0.11 0.28 ± 0.09 Lead NS NS NS 0.54 ± 0.02 0.54 ± 0.03 Zinc NS NS NS 33.74 ± 1.91 37.57 ± 2.65 a Group means have been pooled across two habitats (riparian and non-riparian) and six survey sites. b SE = standard error. c NS = not significant.

Part 2 Landscape protection and restoration 53 Ordination indicated that axes I–III

Axis III [16%] ( N, K, Al, Cu,4 Fe, Co) (with 25%, 22% and 16% explanatory power, respectively) are needed to explain differences in leaf chemistry of 2 the investigated species (Figure 23.2). However, the major difference between cat’s claw creeper and other vines lies 0 in the combination of axes II and III. Phosphorus and molybdenum (on axis –2 II) and nitrogen, potassium, aluminium, copper, iron and cobalt ions (on axis –4 III) were the major driver variables. –6 6 Interestingly, leaf chemistry of the Axis I–4 [25%] (Ca, Mg, Na, S, Zn) 4 native P. straminea is the closest to cat’s –2 2 claw creeper of all non-invasive species 0 0 tested, confirming the often-observed overabundance of this native vine in 2 –2 Species remnant vegetation and the notion that the Axis II [22%] (P, Mo) cat’s claw creeper species has to be managed in its own right. 4 –4 P. suberosa P. straminea Funding in 2010–11 S. australis Queensland Government Figure 23.2 Ordination on three principal axes of leaf chemistry in cat’s claw creeper and Collaborators co-occurring introduced non-invasive (P. suberosa) and native (P. straminea and S. australis) species: each data point represents a single plant; arrows indicate direction and magnitude of • Dr Tanya Scharaschkin (Queensland leaf traits driving each axis; percentages refer to data variation captured by each axis University of Technology, Science and Engineering Faculty) More information Osunkoya, OO, Pyle, K, Scharaschkin, T & Dhileepan, K 2009, ‘What lies • Dr Alan Andersen (CSIRO Ecosystem beneath? The pattern and abundance Sciences, Darwin) Key publications Osunkoya, OO, Polo, C & Andersen, AN of the subterranean tuber bank of 2011, ‘Invasion impacts on biodiversity: the invasive liana cat’s claw creeper, responses of ant communities to Macfadyena unguis-cati (Bignoniaceae)’, infestation by cat’s claw creeper, Australian Journal of Botany 57(2): Macfadyena unguis-cati (Bignoniaceae) 132–8. in subtropical Australia’, Biological For further information on this research Invasions 13(10): 2289–302. project and access to key publications, Osunkoya, OO & Perrett, C 2011, visit the invasive plant and animal science ‘Lantana camara L. (Verbenaceae) pages on the Biosecurity Queensland invasion effects on soil physicochemical website at www.biosecurity.qld.gov.au properties’, Biology and Fertility of Soils 47(3): 349–55.

54 Technical highlights: research projects 2010–11 24. Cabomba (Cabomba caroliniana) ecology

Project dates conditions for cabomba, there is little Cabomba establishment— scientific data available about the October 2010 – July 2012 substrate types ecology of cabomba, both from its native We establish 16 outdoor mesocosms and introduced range. Project leader (each 800 L) with a 10 cm substrate Dr Tobias Bickel Such knowledge is necessary to predict layer comprising a lower 5 cm layer of Ecosciences Precinct in which habitats cabomba is likely to substrate mixed with topsoil and an upper Tel: (07) 3255 4476 establish. This will allow concentrated 5 cm layer of clean substrate to prevent Email: [email protected] monitoring in areas deemed high risk, nutrient leaching. There are four substrate and therefore will improve the likelihood treatments (four replicates each): sand Other staff in 2010–11 of detecting cabomba infestations in (< 0.5 mm), fine gravel (< 5 mm), coarse early stages, when successful removal gravel (< 20 mm) and cobbles (> 100 mm). Cameron Clark is viable. A better understanding of The tanks are filled with rainwater and Objectives cabomba ecology is also crucial for seeded with 75 g wet mass of cabomba mitigating ecological and economic material consisting of approximately 100 • Research the ecological habitat impacts of this serious pest. stem fragments (a mix of single-node requirements and competitiveness of fragments and longer stem pieces). After cabomba. Methods 8 months we drain the mesocosms and • Investigate the regeneration ability measure the total cabomba wet mass and Cabomba regeneration— stem density. Water quality is monitored and establishment of cabomba nutrients in solution fragments to predict dispersal. throughout the experiment. Cabomba fragments (each consisting Rationale of a stem piece with a single node and Cabomba regeneration— one leaf pair) are incubated in shallow substrate quality and pH Cabomba (Cabomba caroliniana) is a plastic containers filled with culture submerged aquatic weed originating from Four 100 L aquariums are filled with a solution. There are five treatments South America. It is a popular aquarium mixture of tap water and distilled water (five replicates each): no nutrients, species and was introduced to the wild in the laboratory and trace elements and trace elements only, low nutrients worldwide through intentional propagation micronutrients are added. In three tanks, (0.5 mg nitrogen L–1), medium nutrients for the aquatic plant trade and disposal of we adjust the pH through CO injection (1 mg nitrogen L–1) and high nutrients 2 surplus aquarium material. Today, it is a to pH 6.5, 7.0 and 8.0; the fourth tank (5 mg nitrogen L–1). serious aquatic weed in many countries, does not receive added CO2 and varies in pH from 7.5 to 8.0. In each tank we plant including Australia, the United States These nutrient concentrations are 10 cm long apical cabomba shoots in little and China. Cabomba was first recorded in representative of trophic statuses of pots filled with four different mixtures of Australia in 1967 and is today naturalised freshwater systems: ultra-oligotrophic sand and topsoil: sand only, 25% topsoil, in several states. It is a WONS and a (trace elements only), oligotrophic (low 50% topsoil and 100% topsoil. There are Class 2 declared weed in Queensland. nutrients), mesotrophic (medium nutrients) three pots of each substrate quality per Even though cultivation and sale of and eutrophic (high nutrients). We inspect tank. Shoot growth and establishment cabomba is now prohibited, the plant is fragments weekly and assess the number of new shoots is then monitored over increasing its range and could potentially of regenerating fragments (new shoot establish in large parts of Australia with 6 weeks, after which we harvest all development from nodes). suitable habitat. Cabomba predominantly cabomba and measure the dry mass of reproduces through vegetative propagules shoots and roots. (stem fragments); viable seeds have so far 70 been observed only in Northern Territory no nutrients trace elements only 60 populations. Cabomba readily spreads low nutrients within catchments through the movement medium nutrients of fragments via water currents, 50 high nutrients particularly in floodwaters. However, 40 humans are the main vector for dispersal between waterbodies, mainly through 30 boating and fishing activities (fouling of fragments in equipment). (%) Regeneration 20

Once established, cabomba has serious 10 environmental and economic impacts and is difficult to control due to limited 0 availability of effective control options. 0 1 2 3 4 5 6 7 8 9 10 While there is general knowledge in Weeks after fragmentation aquarium literature about culturing Figure 24.1 Regeneration of cabomba fragments (see Photo 24.2) as affected by nutrient availability

Part 2 Landscape protection and restoration 55 Progress in 2010–11 Cabomba regeneration— nutrients in solution Cabomba regenerated readily from single-node fragments, independent of nutrient availability in the culture solution (Figure 24.1 and Photo 24.1). Only 4 weeks after fragmentation, about 50% of all fragments grew new shoots. Towards the end of the experiment, there was a decrease in cabomba health in the high and medium nutrient concentrations due to excessive algal growth. Overall this means that cabomba does not depend on nutrients available in the water to regenerate, but is probably able to use internally Photo 24.1 Cabomba fragments stored nutrients for initial growth. Consequently, cabomba would be able to establish from single-node fragments 1200 even in habitats with low nutrient standard deviation availability (oligotrophic systems). 1000 ) Cabomba establishment— –2 800 substrate types

The experiment was conducted from 600 March to December 2010. There was a pronounced difference between cabomba 400 establishment in the mesocosms with (g m mass wet Shoot different substrate types. Cabomba 200 clearly prefers fine substrates (sand) and did not establish well even in fairly 0 fine gravel (Figure 24.2). However, there sand fine gravel coarse gravel cobble was a fairly high variability in cabomba Substrate type establishment, even within the same Figure 24.2 Cabomba establishment and growth, measured as shoot wet mass, as affected by substrate treatment, indicating that other substrate type factors are also important. Nevertheless, while two of the sand mesocosms were completely filled with cabomba after 8 months (Photo 24.2), there was very little establishment of shoots in other substrate types. This clearly shows that cabomba establishment is limited to suitable substrates.

Photo 24.2 A fine sand mesocosm with strong cabomba growth after 8 months

56 Technical highlights: research projects 2010–11 Cabomba regeneration— 0.6 all pH treatments substrate quality and pH pH 6.5 pH 7 0.5 Cabomba fragments grew best in pH 8 substrates with a low organic content, pH variable 0.4 but also did well in substrates with a standard deviation large proportion of topsoil (Figure 24.3). 0.3 The sand-only treatment, which lacked organic content, was the only 0.2 substrate that did not support good cabomba growth. However, growth in (g) dry mass shoot Average substrates was also influenced by the 0.1 pH (Figure 24.3 and Photo 24.3). The best growth was achieved at pH 6.5 in 0 0 25 50 100 the substrate with 25% topsoil. This Substrate quality (percentage topsoil) shows that cabomba prefers slightly acidic conditions and the pH of the Figure 24.3 Cabomba establishment and growth, measured as shoot dry mass, as affected by substrate quality and pH of the water water is an overriding factor in overall growth performance. This experiment also indicates that cabomba depends upon organic content in the substrate for healthy growth. One possible reason for this could be that cabomba might not be able to take up all its nutrient requirements from water alone, but also depends on nutrient supply through the substrate.

Conclusions Cabomba is able to regenerate successfully even from single-node fragments. After only 4 weeks, up to 50% of all propagules produced new shoots. Regeneration does not appear to depend on nutrient concentrations in the water. The pH of the water seems to be an overall limiting factor, with the Photo 24.3 Cabomba shoots growing in an experimental aquarium to investigate the influence of substrate quality on growth plant preferring water that is slightly acidic. Cabomba appears to be able to establish Collaborators even in nutrient-poor waters as long as More information nutrients are available in the substrate. • Brisbane City Council For further information on this • CSIRO research project, visit the invasive Funding in 2010–11 plant and animal science pages on the • Seqwater Land Protection Fund ($120 000) Biosecurity Queensland website at • National Aquatic Weed Management www.biosecurity.qld.gov.au Group • Noosa and District Landcare • Department of Primary Industries, Victoria

Part 2 Landscape protection and restoration 57