Stephen Vlahos, Belinda B Bastow and Gerry A Rayner

Worsley Alumina Pty Ltd, Boddington Bauxite Mine, PO Box 50 Boddington WA 6390


Worsley Alumina Pty Ltd (Worsley) manages the Boddington Bauxite Mine located on the eastern edge of the Darling Range in the northern Jarrah forest of Western Australia. Crushed bauxite is transported 51 km by conveyor to the Worsley refinery. Current annual production is 7 Mt of bauxite requiring approximately 80 ha to be mined and rehabilitated each year. With expansion of refinery capacity to process 12 Mt of bauxite, mining and rehabilitation will increase to approximately 140 ha in the year 2000.

Prior to the commencement of mining operations a trial rehabilitation experiment in 1981 demonstrated that the then current rehabilitation techniques could establish a forest vegetation in rehabilitated bauxite pits. The first operational rehabilitation was carried out in 1986. Since the rehabilitation trial incremental improvements in forest rehabilitation techniques have been implemented based on the results of ongoing monitoring and research conducted by both Worsley Alumina and within the mining industry as a whole.

This paper outlines the evolution of rehabilitation techniques at the Boddington Bauxite Mine.


rehabilitation, bauxite, jarrah forest, mining



Stephen Vlahos, Belinda B Bastow, and Gerry A Rayner


The Boddington Bauxite Mine (BBM) is located on the eastern edge of the Darling Range, 15 km south of the town of Boddington and approximately 130 km south-east of Perth, Western Australia (Figure 1). Project construction commenced in 1981 with commissioning of the mine in 1983. The BBM operation involves the mining and crushing of bauxite for conveying to the Worsley Refinery 51 km to the south-west.

Figure 1

Location of the Boddington Bauxite Mine in south-west Western Australia.

Exploration of the eastern part of the Darling Range was undertaken during the early 1960s and 1970s following identification and development of the bauxite resources on the western edge of the Darling Range by Alcoa of Australia. Exploration of a large area over a number of years identified a core bauxite resource, the primary bauxite area (PBA) in the Boddington area. The primary bauxite resources are located in both state forest and private land. Land uses within the area include farming, forestry, recreation and gold mining.


1.1 Climate

The area has a typical Mediterranean climate with hot dry summers and cool wet winters. The BBM has an average annual rainfall of 740 mm: 50% of rainfall occurs in winter (June-August period) with less than 6% in the summer (December–February period). Average daily temperatures range from a maximum of 32 C in January to a minimum of 5 C in July
(Figure 2.). Maximum temperatures exceed 40 C in summer and fall below 0 C in winter. It is not uncommon for the area to experience consecutive days of frost during winter.

Figure 2

Average monthly climate data for the Boddington Bauxite Mine

1.2 Vegetation

The vegetation on the Darling Plateau is a dry sclerophyll forest dominated by Eucalyptus marginata (jarrah) with varying proportions of Corymbia calophylla (marri) and Allocasuarina fraseriana. E. wandoo (wandoo) forest and woodlands occur in the clay soil valleys and on some upper slopes throughout the eastern edge of the Darling Plateau.

The Jarrah forest mid-storey is predominated composed of trees from the Proteaceae family (including Banksia grandis, Persoonia longifolia, Dryandra sessilis, and Hakea prostrata), Allocasuarina and Melaleuca species. Shrub and herb species from Proteaceae, Papilionanceae, Myrtaceae, Mimosaceae and Asteraceae families constitute a major proportion of the understorey vegetation. Within the current mining envelope the vegetation comprises approximately 400 species from 180 genera. The distribution of these species is associated with landform and soil type (Worsley Alumina, 1985).

1.3 Geomorphology and Soils

The BBM operation lies on the eastern edge of the Darling Plateau. The ancient plateau soils were primarily formed by the in situ weathering of the igneous rock. Laterisation of the region is generally considered to have occurred during the Tertiary Period. Approximately 80% of the bauxite mined to date has been derived from a mafic volcanic (greenstone) parent material that has a high iron and generally low silica content (Ball & Gilkes, 1985). The greenstone-derived bauxite is limited in distribution and is surrounded by the more typical Darling Range bauxite/laterite derived from granite.

The landscape of the Primary Bauxite Area is undulating with uplands ranging from 280-400m above sea level with gentle-to-steep slopes to the valley floors 50-100m below. The landforms and soils in the region are described in detail by McArthur et. al. (1977). Bauxite is contained primarily within the gently undulating laterised uplands with gravelly sands and sheets of duricrust of the Dwellingup landform.

The bauxite is overlain by an average 0.8 m of topsoil and gravel overburden although in some areas the pisolithic gravel is up to 4m in depth. The underlying bauxite horizon is an average of 6 m thick with the top 50% of the total bauxite profile composed of a strong iron-cemented hardcap. The remaining bauxite is friable and earthy. In some deposits derived from the greenstone belt the bauxite profile is over 20 m thick. A series of kaolinitic and montmorillonitic clays up to 30 m thick underlie the bauxite.

1.4 Exploration and Mining Operations

Exploration drilling is conducted on a 100 m or greater square grid. Areas shown to be prospective are re-drilled on a 50 m grid to provide mine-planning data, which is used to identify prospective bauxite pods. These pods are then drilled on a 25 m grid to enable detailed planning and mining control.

The Department of Conservation and Land Management (CALM) is given at least 6 months’ notice prior to the clearing of crown land scheduled for mining. The Department arranges for the harvesting of merchantable forest products including sawlogs, fencing timber and dry and green firewood. Grass trees (Xanthorrhoea) have also been collected for garden landscaping. Current estimates are that 60% of the standing forest biomass is harvested. A further 10% of the forest residue is currently retained for use in the development of fauna habitats during rehabilitation and the remaining forest residue is windrowed and burnt.

On completion of clearing, the topsoil and gravel overburden are separately stripped and directly return to nearby areas undergoing rehabilitation or stockpiled for later use. Once exposed the hardcap layer is either, drilled and blasted or broken by ripping with a very large bulldozer (Komatsu 575). The bauxite is loaded into 83 t haul trucks using front-end loaders or excavators for transport to the crushing facility. The two stage crushing facility reduces the bauxite to less than 35 mm diameter. Once crushed, the bauxite is stockpile on a surgepile and fed onto the overland conveyor for transport to the refinery.


The first operational rehabilitation was undertaken in 1986 (an area of approximately 40 ha). The area of annual mine rehabilitation has progressively increased over time and since 1994 an average of 70 ha has been rehabilitated each year. To date a total of 750 ha of forest has been rehabilitated on crown and private land.

2.1 Rehabilitation Prescription

The rehabilitation prescription was first developed in 1984 in consultation with the then Forestry Department (now CALM). CALM is the government agency responsible for management of State forests. The primary objective for the rehabilitation of forests on crown land is to regenerate a self-sustaining forest ecosystem that is compatible with surrounding forest. Specifically, Worsley aims to maintain the recreational, conservation, landscape and hydrological values of the north-eastern Jarrah forest while minimising the impact on adjacent unmined areas.

Worsley’s rehabilitation prescription has evolved through a continuous improvement process which has been ongoing since 1986. Rehabilitation monitoring and research findings together with technological improvements are reviewed annually in consultation with CALM and continue to be incorporated into rehabilitation operations. An overview of the rehabilitation program and improvements undertaken over the past 12 years is provided below.

2.2 Landform Reconstruction

Following mining the former bauxite pit is shaped to blend with the surrounding landscape to facilitate future land uses. In doing so the earthworks are designed to create a safe and stable landform. This includes reducing the slope of pit batters to less than 33% and the development of surface water exit points to prevent ponding.

Pit floors are ripped on the contour to a depth of 1.5 m at a spacing of 1 m to reduce clay compaction which might otherwise inhibit water infiltration and root penetration. In early rehabilitation operations a straight ripper was utilised, however more recently this operation is conducted by cross ripping or using a ‘winged’ ripper. Cross ripping or winged ripping greatly improve the breakup of compaction compared with the conventional ripper.

Gravel and topsoil are returned either directly from areas being developed for mining or from stockpiles. Placement of the overburden further enhances shaping and blending of the pit into the surrounding topography. Once the topsoil is returned to the pit the area is scarified on contour to a depth of 30-40 cm. This reduces any surface compaction formed by machinery during the return of overburden. Scarification also improves surface infiltration reducing the potential for erosion resulting from run-off.

2.3 Fauna Habitats

In the early rehabilitation the focus on habitat restoration was on the development of niches for invertebrates and reptiles. The approach was one of providing habitats through vegetation development and the return of forest litter with topsoil. At that time no specific provision was made to introduce habitats for vertebrate fauna until the lower order food chain could be established.

Since 1994, provision for specific vertebrate habitats have been introduced into BBM rehabilitated areas. During forest harvesting and clearing operations logs, stumps and forest residue are stockpiled for subsequent inclusion in the rehabilitation. Following the return of topsoil, logs and forest residue are scattered throughout the rehabilitation area. Piles of partially buried logs and stumps are also created providing potential habitats for burrowing animals including the Chuditch (Dasyurus geoffroii), Echidna (Tachyglossus aculeatus) and Monitor lizards. To encourage arboreal (tree dwelling) marsupials including the Brushtail Phascogale (Phascogale tapoatafa) nesting boxes have been attached to trees within the oldest areas of rehabilitation. Use of these artificial habitats by native marsupials has been confirmed by the identification of hair collected from nesting material within the boxes.

Systematic monitoring of vegetation in the earlier areas of rehabilitation together with surrounding unmined forest is used to determine fauna utilisation and evaluate further habitat requirements.

2.4 Revegetation

Once the rehabilitated earthworks have been completed the pits are broadcast seeded with a range of locally collected understorey and tree species. Sowing was undertaken during the wet months of May to July. Research by Alcoa has indicated that there may be benefits to establishment from sowing earlier in the year rather than during the winter months (Ward et. al. 1996). Trials conducted at the BBM to further develop this premise indicated that areas sown during April and May demonstrated improved plant establishment or growth (as expressed by foliar cover) than areas sown in June and July (Figure 3).

Figure 3

The Effect of Time of Sowing on Plant Density

Seeding activities have historically been undertaken during June and July. As a result of the recent field trials, the timing of this activity has been under review with the view of exploring opportunities to further improve rehabilitation at BBM. Another operational change which has improved the establishment of vegetation in rehabilitation has been to broadcast the seed immediately after the soil is scarified. This improved vegetation establishment is attributed to incorporation of the seed onto the soil before the development of surface crusts (Ward et. al. 1996).

The composition of the original seed mix used in early pit rehabilitation at BBM comprised the dominant understorey species identified within the Saddleback Timber Reserve (Table 1). These species constituted approximately 75% of the understorey base mix.

Table 1

Dominant Understorey Species included in the Original BBM Rehabilitation Seed Mix

Acacia alata

Acacia celastrifolia

Acacia drummondii

Acacia pulchella

Boronia fastigiata

Bossiaea ornata

Daviesia decurrens

Daviesia rhombifolia

Dryandra lindleyana *

Hakea lissocarpa

Kennedia cocinea

Kennedia prostrata

Macrozamia reidlei

Xanthorrhoea pressii

Note: * previously referred to as Dryandra nivea

The diversity of the seed mix has been progressively increased since rehabilitation commenced (Figure 4). These improvements are a result of ongoing monitoring and research, notably through collaborative investigation with Alcoa into recalcitrant species. During 1998 over 120 species were included in the seed mix compared with approximately 55 in original 1986 rehabilitation prescription. Ongoing research continues to focus on collecting the seed of species not returning to rehabilitation. Emphasis on continuous improvement of the rehabilitation prescription has also resulted in a reduction in the proportion of dominant understorey species in the seed mix.

Figure 4

Number of Species Included in the Seed Mixes each Year

Two specific changes to the composition of the BBM seed mix have occurred since the original prescription. During 1993 A. celastrifolia, a ubiquitous legume shrub in the eastern edge of the Jarrah forest, was removed. Its initial inclusion in the rehabilitation seed mix combined with the large seed bank in the returned topsoil resulted in the shrubs dominating the vegetation. Soil conditions in the rehabilitation possibly favour the shrub where high phosphorus levels may increase the competitive advantage over the non-legume species. Although A. celastrifolia has been excluded from the seed mix it still comprises approximately 5% of the total plant density in the establishing vegetation. The proportion of Kennedia species in the seed mix has also been reduced, as these species were found to smother other plants.

In addition to understorey species, seedlings from a range of tree species are planted in the BBM mine rehabilitation during June and July. During the early development of the rehabilitation prescription the fungal disease jarrah dieback (Phythothora cinnamomi) was considered a threat to the persistence of both the jarrah forest and the rehabilitated vegetation. As a result eastern Australian eucalyptus with resistance to dieback were included in the rehabilitation as a safeguard to reinstating forest vegetation. These species included E.mellidora, E. resinifera and E. sideroxyilan. During the period 1988-91 these species were progressively reduced in the seedling mix and since 1992 only local provenance tree seed has been used. The perceived threat of dieback to the revegetation has been allayed with the successful establishment and persistence in rehabilitation of local species by both Worsley and Alcoa.

The impact of the non-native trees on the rehabilitation and surrounding forest vegetation does not appear to be significant. Monitoring of the rehabilitation vegetation in which these species were planted indicates that no volunteer seedlings have been produced. Since 1997 a program has been in place at BBM to remove the non-native trees from the rehabilitation by a selective herbicide treatment. Monitoring indicates recruitment of jarrah and marri is occurring in these 8-12 year old rehabilitation areas.

The establishment of trees in BBM rehabilitation was originally undertaken by transplanting nursery raised seedlings and supplemented by direct seeding. Depending on the position in the landscape a mix of species was planted at a density of 500 tree per ha. Initial prescriptions included the broadcast seeding a mixture of E. marginata, E. wandoo and Cormybia calophylla. During the evolution of the BBM prescription direct seeding was seen as the preferred method of establishing trees, offering advantages in operations and, more importantly developing a better form of jarrah tree. Jarrah trees developing from planted seedlings tend to have a higher occurrence of multiple stems than trees establishing from seed.

Direct seeding trials conducted at the BBM demonstrated that trees can be established solely by broadcasting seed. In a trial sown in 1988 total tree numbers increased from 344 per ha at 12 months of age to 770 per ha after 9 years (Figure 5). Progressive changes to techniques for tree establishment has reduced the reliance on planting seedlings. With the reduction of planting to 250 seedlings per ha there has been an increase in tree species sowing rates. Not all rehabilitated areas are however, planted to seedlings. Seedlings are specifically planted on areas of high visual exposure and steeper slopes.


Figure 5

Increase in the Density of Tree Seedlings in the Direct Seeding Trial Sown in 1988.

Both seasonal conditions and seed quality are factors contributing to the variability of tree establishment each year. Testing of Jarrah seed has found germination varies from 3 to 80 % between seed lots and years. Maintenance of the seed store with collections from several years has reduced the effect of a single year’s seed quality on the establishing tree density.

2.5 Fertilising

Nutrients within the rehabilitated pits are supplemented with 450 kg/ha of superphosphate (plus trace elements of copper, molybdenum and zinc) broadcast by helicopter during August. Additional nutrients are supplied to each tree with a 200 g pellet of DAP (di-ammonium phosphate) buried adjacent to each seedling at the time of planting.



A systematic monitoring program to assess the establishment of trees and understorey plants is undertaken during autumn (March- April) following seeding. Where rehabilitation does not meet the required standard supplementary revegetation is undertaken.

In developing the rehabilitation prescription, performance indicators have been identified to assess the establishment of vegetation indicative of a sustainable forest ecosystem. Initial success criteria for establishment included 80% survival of transplanted seedlings and an average density of 600 tree per ha. A second criterion was for the understorey density to average 1 plant/m2. The criteria for understorey success is regularly reviewed in consultation with CALM and now comprise 2 plants/m2 of which 1 must be a legume and 1 a non-legume. To date all years of rehabilitation have achieved the required standard (Figure 6).

Figure 6

The Density of Understorey Plants in the First Autumn after Seeding within the Rehabilitation Vegetation

Further flora monitoring of the rehabilitation is undertaken to assess development of the tree and understorey vegetation. The current monitoring program assesses permanent plots in rehabilitation at 1, 4, 7 and 10 years of age. Assessment of tree species is conducted to determine tree density (Figure 7 ), height and the occurence of multi-stemming. Understorey vegetation is also assessed; species and plant density is measured and compared with reference plots in the surrounding forest. The results for 1997 are shown in Table 2.


Figure 7

The Average Density of Trees with Increasing Age of Rehabilitation


Table 2

Average Species Richness and Cover, and Density in Rehabilitation

Age and year of rehabilitation

Species richness per 80 m2

Foliar cover (%)

Plant density (number/m2)

1 (1996)




4 (1993)




7 (1990)




10 (1987)




Forest plots





Worsley’s mine rehabilitation research program is based on results from the comprehensive rehabilitation monitoring program together with ongoing technical developments within industry.

This extensive research program covers numerous aspects of rehabilitation. Investigations are undertaken internally, through joint projects conducted in conjunction with other companies (such as Alcoa) and with research institutions. Current studies include:

Worsley’s commitment to continuous improvement and objective of returning a stable and self-sustaining ecosystem means that the rehabilitation research and investigations initiated at BBM in 1986 will continue throughout the life of the project.


Ball PJ and Gilkes RJ (1985). The Mt. Saddleback bauxite deposit south Western Australia. Proceedings of the 3rd International Conference on Laterisation Processes. Japan, Tokyo.

McArthur WM, Churchward HM and Hick PT (1977). Landforms and soils of the Murry River catchment area of Western Australia. Land Resources Management Series No.3, CSIRO.

Ward SC, Koch JM and Ainsworth GL. (1996). The effect of timing of rehabilitation procedures on the establishment of a jarrah forest after bauxite mining. Restoration Ecology, 4, pp 19-24.

Worsley Alumina Pty Ltd (1985). Worsley Alumina Project. Flora and Fauna Studies, Phase Two, Worsley Alumina Pty Ltd, Perth.