Insect resistance and silvicultural control of the shoot borer, Hypsipyla robusta, feeding on species of Meliaceae in Southeast Asia and Australia

Lao PDR, Malaysia, Philippines, Thailand, Vietnam

Dr Rob Floyd
Phone: 02 6246 4155
Fax: 02 6246 4089
Email: Rob.Floyd@csiro.au

• National Agriculture and Forestry Research Institute, Laos
• Forest Research Institute of Malaysia, Malaysia
• Royal Forest Department, Thailand
• Department of Environment and Natural Resources, Philippines
• Queensland Forest Research Institute, Australia
• Forest Science Institute of Vietnam, Vietnam

$1,145,010

01/01/1999 - 31/12/2002

01/01/2003 - 31/12/2003

Dr John Fryer

Hypsipyla robusta moth is a highly damaging forest pest in terms of both level of attack per individual tree and percentage of trees attacked in each plantation. Hypsipyla lays its eggs on the shoots of young trees of the family Meliaceae, in particular the sub-family Swieteniodeae. Several valuable tropical timber trees belong to this group - for example, Toona ciliata (the Australian red cedar), Chukrasia tabularis, and the mahoganies (Swietenia species and Khaya species).
The larvae of the moth tunnel within the top shoots of young trees, eventually killing the shoots. The moth pest is widely distributed in the tropics and severely limits the commercial growing of susceptible trees. In some places, this pest is starting to be successfully managed, but it still remains a problem in India and Sri Lanka, Southeast Asia, West Africa, Central America, Australia and the Pacific.
Some individual trees appear to be more resistant to H. robusta than the norm. Their genotypes need investigating as part of the effort to determine the mechanism of resistance. However, use of genetically resistant types would only be part of a broader approach to control the problem, since resistance may be only partial.
This project linked with the AusAID-funded SPRIG project, which is making an extensive collection of T. ciliata seed, and also with ACIAR project FST/1996/005 on the domestication of C. tabularis and T. ciliata.

The project studied resistance to the shoot borer pest in a family of high-value tropical timber trees. Its aims were to identify resistant tree genotypes, discover the mechanisms of resistance and then trial mixed-species plantings.

The project team worked with forestry departments in Lao PDR, Malaysia, Philippines, Thailand and Vietnam. The first objective was to identify resistant genotypes of Swietenioideae species by trials in the partner countries. In each of the trials the team studied the mode of resistance (for example, avoidance by the pest, tolerance by the tree shoot) with detailed field observations. The CSIRO Division of Entomology analysed data from these trials, and then measured the physical and chemical attributes of seedlots showing a range of different resistances.
In another experiment, the team assessed the effects of companion planting (two non-target species growing in parallel with the susceptible tree) and under-planting (a range of non-target species and different ages planted under existing susceptible trees). Results helped determine how to manage the pest problem by means of diversity in plantations.
In another strand, the scientists investigated the possibility that early fertiliser will cause rapid growth to a stage beyond which the pest attack is significant; they also tested the effectiveness of insecticide dosing at critical times of tree growth. Finally, in collaboration with the project team from FST/1996/005, a survey of H. robusta incidence and damage on C. tabularis was conducted across the native range of the tree in Vietnam, Lao PDR and Thailand.

Hypsipyla robusta is a highly damaging forest pest in terms of both level of attack per individual tree and percentage of trees attacked in each plantation. The project scientists evaluated a wide range of genotypes for genetic resistance to Hypsipyla, and investigated whether silvicultural manipulations could ameliorate the problem.
Four Toona ciliata and 11 Cedrela tabularis genetic resource trials were established and monitored. They found no outstanding seedlots of Toona ciliata in terms of resistance to Hypsipyla but there is some possibility that individual genotypes may prove to have some resistance. Other trials and experiments focused on issues including pruning, fertilising, insecticides, shade effects, and host selection.
Height: The dominant pattern in all trials was a positive relationship between tree height and Hypsipyla damage. If one was to select for fast-growing trees one would unwittingly select for trees that attract more insect damage. The focus for improvement must be on selecting trees that are relatively little damaged in spite of good growth, or that maintain good form in spite of Hypsipyla damage.
Species differences: Khaya senegalensis, Cedrela odorata and C. tabularis performed well relative to the average T. ciliata. In general these three species grew longer boles and received less Hypsipyla damage. K. senegalensis was the least frequently damaged species. The relative merit of each species depended on the site.
Pruning: Pruning trials on C. tabularis and T. ciliata found no advantage in form or growth of trees.
Shade: Growth of T. ciliata was examined under an overstorey and in a forest gap. These trees showed performance that exceeded any open-planted T. ciliata. Laboratory experiments support this result, indicating Hypsipyla robusta prefers to lay eggs on leaves grown at high light. These experiments suggest the most promising approach to T. ciliata silviculture is management of a light environment that optimises the balance between tree growth and insect damage.
Plant chemistry: Scientists found that leaf chemistry predicts variation in the H. robusta damage of T. ciliata. Gas chromatography of T. ciliata leaf extracts shows that trees with low damage have more of the compound bicyclo-elemene. They also found differences comparing leaves from high and low light, correlated with the shading effect on oviposition. Together these results indicate that variation in H. robusta attack is driven by plant chemistry. Damage reduction requires environments that make the trees constitutively less attractive to H. robusta.
Genetic improvement: In the fastest growing C. tabularis trial the best seedlot had mean bole lengths approaching 4 m after 36 months. Seedlot 20204 was a consistently good across many sites, ranking among the best with regards bole length, frequency of Hypsipyla damage and branching. In contrast, Hypsipyla damage was so intense in the T. ciliata genetic resource trials that no seedlots and very few individual trees looked 'forestry-ready'. Nevertheless there were significant effects of seedlot on important traits. For example, seedlot 109 was significantly better than most seedlots with regards bole length, frequency of Hypsipyla damage, and branching. The presence of seedlot effects on plant traits suggests selection or cloning has the potential to produce better trees.
Also, some seedlots were ranked in the top ten across several of the countries and these, plus a small number of well-performed individual trees could form the basis of further research and demonstration trials. Some seedlots of non-Toona species such as Chukrasia and Khaya in some countries have potential for promotion for future plantings because of their stronger growth, lower attack by Hypsipyla and their ability, particularly in the case of Chukrasia, to form a vigorous new terminal shoot with a high angle of recovery.