Biological control of Chromolaena odorata in Indonesia and Papua New Guinea

• Universitas Nusa Cendana, Indonesia
• SEAMEO Regional Centre for Tropical Biology, Indonesia
• Department of Agriculture and Livestock, Papua New Guinea
• Parks and Wildlife Commission of the Northern Territory, Australia
• Gadjah Mada University, Indonesia
• Philippine Coconut Authority, Philippines
• Indonesian Oil Palm Research Institute, Indonesia
• Centre de Cooperation Internationale en Recherche Agronomique pour le Developpement, Indonesia
• Oil Palm Research Association, Papua New Guinea

Siam weed (Chromolaena odorata) is a serious weed of tropical pastures and a threat to national parks and other biodiversity conservation areas. It is a prolific seed-producer and is usually toxic to livestock. It has quickly spread from its original home in the West Indies to large areas of the wet tropics of Africa and Asia. It reached Timor in the mid-1970s and invaded parts of Irian Jaya, New Britain and possibly the Solomon Islands. Siam weed is now considered the major weed threat to Australia. Efforts are still under way to contain an infestation originally discovered in Tully, Queensland, in 1994.

There are many natural enemies of this weed and biological control has long been touted as an option. By 1992, only two agents had been tried - a seed-eating weevil (Apion brunneonigrum) and a leaf-eating moth (Pareuchaetes pseudoinsulata), which were both tried in South Africa. ACIAR project CS2/1991/010 released the moth into oil palm and pastoral areas in Java, Sumatra and Timor. The moth successfully established in several areas in Sumatra, where there is now effective control of the weed. However, the agent did not persist in Java or Timor.

Project CS2/1991/010 also introduced a new control agent, the gallfly Procecidochares connexa, from South America, which had not been tried or host-tested anywhere else in the world. The first releases were made in Indonesia in July 1995; it was also intended for release in the Philippines but government permission was not granted during the lifetime of the project. However, permission came through in February 1997 and this agent was released in this project.

The aim of this project is to enhance the biological control of Siam weed (Chromolaena odorata) in Indonesia, Papua New Guinea and the Philippines. The work continues that begun in a previous ACIAR project (CS2/91/10), which released control agents in Indonesia. The current project intends to carry out further releases, conduct field monitoring of their effectiveness and introduce additional control agents.

Mass-rearing and release of the moth and the stem-galling fly will take place in Indonesia and PNG. A new control agent, a leaf-feeding butterfly (Actinote anteas), will be host-tested in Indonesia. Application will then be made for a release permit.

An entomologist in South America will survey insects feeding on Siam weed and related plants. Promising insects will be studied and likely control agents chosen. Host-testing of these will be carried out in Indonesia, once import permits are granted, to ensure that no other plants are attacked.

The next step will be mass-rearing and releases of new agents at sites in Java and Timor. The work in Timor will be in collaboration with staff from the Northern Territory Parks and Wildlife Commission.

Meanwhile, host-testing of the gall fly, P. connexa, will take place in the Philippines and, following the granting of another permit, field releases will start.

Weed population densities at selected sites will be measured before and after release of control agents. Agent populations will also be monitored to estimate the rate of spread and of population increase, and to identify possible parasites and predators of the control agents.

This project was active from 1997 to 2007. During that time attempts were made at a multitude of sites in Indonesia and PNG to establish different potential control agents and monitor their effectiveness. No satisfactory outcomes were achieved in Philippines where, although a permit was eventually granted to test the gallfly, no permit could be obtained for its release, so further efforts with biocontrol were terminated.

In the period 1997-2001 the project achieved successful control of chromolaena in North Sumatra and increasing control in West Java and other areas. In estate crops such as rubber, oil palm, or coconut, there is a significant economic impact in terms of reduced herbicide use and reduced slashing rounds in the establishment phase of plantations.

However, the real economic impact is where crops are combined with livestock, and the cattle fed on plants growing in the fallow or off-season land, either by direct grazing or by cut-and-carry methods. Chromolaena is poisonous to cattle and is not grazed, but out-competes and replaces grasses and fodder legumes. Thus land invaded by the weed is not available for livestock. This has a very significant impact, especially in the eastern islands where livestock raising is an important part of the local economy. In the absence of chromolaena, other invasive weeds such as lantana may take over, but these can be controlled more easily as they are less aggressive, particularly where fire is used.

In much of Indonesia and Papua New Guinea, chromolaena was not found to be a significant environmental problem, as it is only invasive in cleared or logged forest. However, where these forests were being preserved as reserves, the invasion of chromolaena along tracks and in clearings worsened the degradation and loss of natural understorey caused by logging and other human activities. Control by the gallfly allowed the natural vegetation to compete.

In natural grasslands such as in Timor and other eastern islands, the environmental impact of chromolaena was much more severe. These grasslands were already degraded through over-grazing, but chromolaena will invade even intact savannahs wherever the rainfall is adequate and completely replace the grass. This has severely impacted on the native vegetation and wildlife as well as destroying the livestock industry. In the period 1997-2001 control by the gallfly was not yet adequate in the eastern islands, but the degree of control increased each year and seed production and seedling growth diminished greatly.

For Australia, the impact was two-fold. First, reduced seed production in the eastern islands of Indonesia and in PNG would reduce the risk of seed being transported to Australia. The second positive outcome was that, should a large infestation of chromolaena be discovered in Australia such that eradication is not practicable, the gallfly would be immediately available for release to provide immediate biocontrol.

The gallfly proved a great success in all countries in which it was released. It has the capacity to locate and kill single plants and would be a worthwhile agent in any country where chromolaena is a problem.
For the period 2002-05 chromolaena was reported at about 500 sites in 13 lowland provinces of PNG. The gallfly was released in all provinces and its presence confirmed at about 400 sites, with about 60 unconfirmed sites. Control of chromolaena was reported at various sites in five provinces, namely Bougainville, Sandaun, New Ireland, and East and West New Britain. The gallfly should be able to spread naturally to most of the remaining sites. The moth Pareuchaetes pseudoinsulata was released in nine provinces but established in only one province, Morobe and is seasonally present at 20 sites.

Field monitoring and observations suggested that the gallfly performs best in shaded or semi-shaded areas, where there is high moisture or rainfall. In dry areas such as the Markham Valley in Morobe and open and exposed areas in West New Britain, the gallfly was slow to build up into damaging populations and chromolaena was not under control, despite its presence for over 5 years. P. pseudoinsulata appeared to prefer drier areas such as the Markham Valley, but given the level of effort required to achieve establishment, this agent is only recommended for release in exceptional circumstances.

A new agent, Calycomyza eupatorivora, was imported during this period, but there were problems with its rearing and release and it failed to establish at any site. This agent was easy to rear and establish in South Africa in Durban, where it quickly spread. However, it is possible that the areas in PNG are too hot for the insect. The project team concluded that this insect is more suited to cooler countries or regions.

In areas where the agents are less effective, land managers will still need to implement some other control measures. However, the time and cost involved should reduce in the presence of the agents.

In PNG the project involved numerous staff from the National Agricultural Research Institute from several regions. In addition, project officers collaborated with staff from government and commercial institutions. This co-operation between the organisations has helped with the success of this project, with regional officers helping to spread the agents around their districts.

A 3-day workshop on the biocontrol of chromolaena was conducted in Lae in 2003. The workshop covered aspects such as biological control of weeds, weed identification, biocontrol agents, host specificity and field release and monitoring of biocontrol agents. As a result of this collaboration, regional officers became more aware of other weeds and took more interest in other projects. As an example, researchers from the Cocoa and Coconut Institute and the Oil Palm Research Association are involved in the ACIAR-funded biocontrol of mikania.