Initial plantings of A. mangium outside its natural distribution range had generally relied on unimproved materials usually from a narrow genetic base. Consequently, the growth obtained was variable and productivity tended to decline over several generations due to genetic erosion (Awang and Bhumibhamon 1993). Elaborate tree improvement activities are now being taken up in many countries where it has been introduced for production of better planting materials with consistent, desirable characteristics.
A. mangium has a fragmented natural distribution stretching from the Moluccas islands in Indonesia to Western Province of Papua New Guinea and northeastern Queensland in Australia. Many provenances highly adapted to their natural habitats have been identified and studies have shown variation among them in all respects (Awang and Bhumibhamon 1993). For example, there are large provenance differences in growth rate, stem straightness and frequency of multiple leaders. International provenance trials were established during the 1980s (Doran and Skelton 1982). One of the international provenance trials are shown in Table 1 (Awang and Taylor 1993).
Table 1. A. mangium provenances in the international provenance trials (Awang and Taylor 1993)
Sl. No. | Provenance location (CSIRO seedlot No.) | Provenance region | Lat. (ºS) | Long. (ºE) | Altitude (m) |
1 | Julatten (12990) | Queensland Cairns Region | 16 34 | 145 35 | 400 |
2 | Daintree (12991) | Queensland Cairns Region | 16 17 | 145 31 | 60 |
3 | Rex Range (12992) | Queensland Cairns Region | 16 30 | 145 32 | 30 |
4 | Claudie River (13229) | Far North Queensland | 12 44 | 143 13 | 60 |
5 | Mission Beach (13230) | Queensland Cairns Region | 17 53 | 146 06 | 5 |
6 | NW of Silkwood (13231) | Queensland Cairns Region | 17 42 | 145 57 | 40 |
7 | Cowley Beach (13232) | Queensland Cairns Region | 17 41 | 146 05 | 5 |
8 | NE Walshs Pyramid (13233) | Queensland Cairns Region | 17 06 | 145 48 | 20 |
9 | E of Cairns (13234) | Queensland Cairns Region | 17 02 | 145 48 | 20 |
10 | Mourilayan Bay (13235) | Queensland CairnsRegion | 17 35 | 146 05 | 20 |
11 | Kurrimine (13236) | Queensland Cairns Region | 17 46 | 146 05 | 10 |
12 | El Arish (13237) | Queensland Cairns Region | 17 50 | 146 01 | 20 |
13 | Mission Beach (13238) | Queensland Cairns Region | 17 56 | 146 02 | 70 |
14 | Tully (13239) | Queensland Cairns Region | 17 55 | 145 52 | 50 |
15 | Cardwell-Ellerbeck (13240) | Queensland Cairns Region | 18 14 | 145 50 | 60 |
16 | Broken Pole Creek (13241) | Queensland Cairns Region | 18 21 | 146 03 | 50 |
17 | Abergowrie S. F (13242) | Queensland Cairns Region | 18 26 | 146 01 | 60 |
18 | Daintree (13279) | Queensland Cairns Region | 16 17 | 145 31 | 60 |
19 | Morehead (13459) | Papua New Guinea | 8 45 | 141 25 | 30 |
20 | Oriomo River (13460) | Papua New Guinea | 8 50 | 143 08 | 10 |
21 | Cassowary Range (13534) | Queensland Cairns Region | 16 32 | 145 25 | 60 |
22 | Piru, Ceram (13621) | Ceram, Indonesia | 3 04 | 128 12 | 150 |
23 | Sidei (13622) | Irian Jaya, Indonesia | 0 46 | 133 34 | 30 |
24 | Mossman (13846) | Queensland Cairns Region | 16 31 | 145 24 | 60 |
Results of these trials were reported by Harwood and Williams (1992). Highly significant provenance differences in growth trait among experimental sites and provenance regions were observed. For example, growth was generally faster at near-equatorial trial sites with mean annual height increment around 3-4 m, and slower at sites further from the equator. Papua New Guinea provenances were consistently the best performers, closely followed by the Claudie River provenance from north Queensland. The slowest growing provenances were from the Maluku province of Indonesia and southern parts of the distribution in Queensland. Other studies have given similar results (Nguyen and Le 1996; Otsamo et al. 1996; Tuomela et al. 1996).
Though variation of wood density among provenances was not significant, the heartwood formation variation was significant among 5-year-old trees of 23 families from 7 seed sources (Bhumington et al. 1992). It is highly heritable trait by the narrow heritability (Awang and Bhumibhamon 1993).
Plus tree selection and progeny testing for selected materials has taken place in several countries where A. mangium is planted on a large scale. Generally, the characteristics considered for plus tree selection are superior height, good diameter at breast height, stem straightness, good branching habit, good self pruning ability, resistance to pest and diseases and wood properties (density).
Since establishment of provenance trials, these countries have started further comprehensive seed collections. For example, Australian and several South East Asian organizations are developing improved breeds from wide bases of the best provenances with low levels of genotype-by-environment interaction at the provenance level (Harwood 1996). A. mangium seed orchards have been established in Australia, India, Indonesia, Malaysia, the Philippines, Taiwan and Thailand (Varghese et al. 1999). Expected gain in seedling volume production in progeny from seed orchards of A. mangium in Indonesia is more than 50-70% compared to local seeds planted (Kurinobu and Nirsatmanto 1996). Because of asynchronous flowering among trees and families in seed orchard (Awang and Bhumibhamon 1993), the seed production of most of the first generation is low. In Fig. 5 the seed orchard in India is shown.
Fig. 5. Acacia mangium seed orchard at Nilambur, Kerala, India.
ТЬ build up a base for clonal forestry programme (Arisman and Havmoller 1994), A. mangium × A. auriculiformis experimental hybrid seed orchards have been established in Indonesia. Hybrid clones with outstanding growth and form selected and propagated by tissue culture are being tested in Vietnam (Le 1996). Vegetative propagation of the hybrid by striking cuttings from coppice shoots is being used in Bangladesh (Banik et al. 1995).
A breeding programme has been initiated by Institute of Forest Genetics and Tree Breeding, India (IFGTB) by obtaining seeds from established seedling seed orchards of Australia, Fiji and Indonesia as well as from identified natural provenances, and raising pedigreed as well as bulked seed orchards after culling inferior families (Varghese et al. 2001).