Botanic Gardens Conservation International
BGCI provides a global voice for all botanic gardens, championing and celebrating their inspiring work. We are the world's largest plant conservation network, open to all. Join us in helping to save the world's threatened plants.

Ecological Restoration of a Cliff Face in Kings Park and Botanic Gardens, Perth, Western Australia

Volume 4 Number 1 - December 2003

Bob Dixon & Peter Moonie

Kings Park and Botanic Garden is located one kilometre from the Central Business District of Perth, the capital city of Western Australia. The park covers an area of 400 hectares, of which two-thirds is bushland (native vegetation).

In 1996, the Botanic Gardens and Parks Authority (BGPA) entered into a partnership with Australian mining company, WMC Resources Ltd, to establish the WMC Centre for Urban Bushland Management. The role of the Centre was to integrate applied research with on-ground operational works to achieve world class ecological restoration. The first project to be undertaken was the restoration of the Kings Park Escarpment, a prominent cliff face of over 70 m in height within the heart of Perth and the site of the State’s first quarry.

The scarp has been subjected to a long history of disturbance which has resulted in extensive areas becoming degraded. Such disturbances have included timber extraction, limestone quarrying, track construction, and planting of exotics for presumed stabilisation and landscape enhancement. As a result, indigenous vegetation cover has been lost, weeds have invaded, incidences of soil slips and rock falls have increased, and biodiversity has declined.

The limestone heath community of the scarp also represents an important bushland remnant with similar community types being poorly represented in the Perth region. Several relic invertebrates of significant conservation value are also restricted in their distribution to the escarpment. There are also a number of priority listed plants within Kings Park that only occur on the escarpment and several species were first plant collected (type collections)  from this site by early botanical collectors.

A research team was appointed as little was known at that time about restoring native vegetation and fauna on such steep slopes covered by highly invasive weeds, such as bridal creeper Asparagus asparagoides. An innovative restoration plan for the scarp was produced which led to the commitment by the State Government of Western Australia to fund the implementation of the restoration plan over a four-year period (Meney, 1999. Mt Eliza escarpment restoration plan. Kings Park and Botanic Garden, West Perth).

Implementation Phase of the Project

Removal of Weeds

Many woody weeds, including non-local native plants, had to be removed (over 3,000 to date). Those removed included large sugar gums Eucalyptus cladocalyx and lemon scented gums Corymbia citriodora, native peppermints Agonis flexuosa, pines mainly Pinus  pinaster, and shrubs such as Geraldton wax Chamelaucium uncinatum. Weeds displace indigenous species, exhibit rapid growth rates, invade relatively closed and intact vegetation communities, promote slope instability and result in the loss of habitat diversity.

A consultant was contracted to conduct a cultural heritage assessment of exotic species on the scarp prior to trees being selected for removal. Those individuals identified as having cultural significance were retained unless they posed a public safety risk. Exotic tree removals were approved provided that neither root matter nor soil was removed off site. A number of individual sugar and lemon scented gums have also been retained in a nominated reference site along Forrest Drive as an educational resource.

Extensive consultation with relevant Aboriginal groups was undertaken in 1998 prior to initial restoration works; this consultation is ongoing.

Prior to the contracts for tree removal being awarded, surveys for rare fauna including several species of trapdoor spider and the scarp snail Bothriembryon indutus were conducted. The populations were mapped and marked to avoid any damage during tree works. All the tree contractors were required to submit fauna protection plans to make sure the fauna was protected. New methods of tree removal were carried out including extensive use of cranes. In one area which was difficult to access, a 400- tonne crane was used to lift a 70-tonne crane further up the scarp to reach trees in remote locations. Most of the trunks and branches were fed directly into a chipper. The chipped material was then trucked directly to degraded sites and used as mulch to suppress weed growth and promote water retention in the soil profile. The use of cranes also assisted in reducing potential damage to the existing indigenous flora and minimising soil slippage and rock fall.

Other weeds, such as the century plant Agave americana with its toxic sap, proved difficult and costly to remove both by hand and machine. Once the leaf blades were physically removed, stumps were painted with Garlon® to kill the plants and reduce suckering. Over 600 cubic metres has been removed to date. The most cost effective method of disposal was tub grinding and the resulting material was used as mulch. Other serious weeds include bridal creeper, bridal veil Asparagus crispus and freesia Freesia aff. leichtlinii. Extensive herbicide trials were conducted to establish an effective means of control. The trials indicated that bridal creeper is best controlled using Brushoff® metsulfuron methyl at the rate of 2.5 g/ha; this also controls many bulbous weeds such as Freesia and Sparaxis. Though the trials indicated it also controlled bridal veil, spot spraying with glyphosate 360 at a rate of 1% is a more reliable method of control.

Perennial veld grass Ehrharta calycina was effectively controlled using a blanket spray of Fusilade® fluazifop-pbutyl at the rate of 4l/ha between June and mid-August when the grass is actively growing and before it flowers. Failure to spray this grass at the correct time results in resprouting from the dormant buds at the base of the plant (Dixon, 1998. Best management practices for the control of perennial veld grass Ehrharta calycina. Managing our bushland, conference proceedings 16-17 Oct. 1998, p 147- 149. Urban Bushland  council, Western Australia Inc.)

Slope Stabilisation

A geotechnical consultant was employed to assess the soil stabilisation risk associated with the proposed works. Provided that exposed slopes >45 degrees were stabilised using previously trialed bioengineering techniques, the risk of soil slippage was considered negligible. Slope stabilisation is achieved by first laying an organic fibre matting to provide interim stabilisation and then revegetating with the appropriate species for long term

The original stabilisation matting used was manufactured from jute; however, it was quickly bleached by sunlight and stood out in the landscape. It also degraded quicker than anticipated. A composite of coconut fibre and goats hair is now used. It is more expensive, but holds its colour and lasts longer. The matting and the metal pins required to bind it to the slope are very expensive, as is the cost of laying the material (Dixon and Moonie, 2001. Erosion control on Kings Park scarp. Western Wildlife (5)4:18-19. Department of Conservation and Land Management, Perth, Western Australia).

A large cliff overhang at the base of the escarpment was at risk of collapsing and required remediation. Geotechnical consultants suggested that an underwall be constructed to mitigate the risk. This included a foundation wall and a limestone rock fascia where deep crevices were left to create native snail habitat, probably the first snail habitat ever created in Australia. The snails ‘over summer’ in the crevices and are therefore protected from  predators such as mice and rats. Unlike the usual garden snails, this species does not feed on live plant material but gains its sustenance from fungi and bacteria in the soil and litter layers.

Excess water runoff from developed areas at the top of the escarpment resulted in a series of erosion gullies along the escarpment. New bioengineering techniques were employed for gully restoration and stabilisation. A series of soakwells have also been placed within a pathway traversing the top of the scarp to intercept excess runoff.


To retain the correct genetic provenance, nearly all seed for revegetation was collected in Kings Park bushland. Seed sourced outside the park is from the same vegetation type and is checked for genetic integrity to make sure it is similar to Kings Park material. Seed is collected by Kings Park staff and volunteers, the location details are recorded using a GPS and seed germinability data are documented for future reference.

Due mainly to propagation difficulties, the initial species list was limited to 35 key species. However, more species have been progressively added to restoration areas as propagation techniques have improved. Up to 60% of the perennial species richness has been returned to some areas. The broad objective for revegetation is to reinstate species richness to foster long-term resilience and sustainability of vegetation as well as faunal communities.

Nearly all greenstock is contract grown in accredited nurseries; however, the Friends of Kings Park growing group and Kings Park nursery staff raise small quantities of species which are difficult to propagate. Over the last 3 years, 350,000 seedlings were planted to replace the 3000 woody weeds that were removed.

The revegetation programme also involves direct seeding where appropriate, many seeds are pre-treated to enhance germination (e.g. legumes are hot water treated and most of the others are smoke treated). In other areas where there is a good seed bank in the soil,  smoke water is applied to stimulate the germination of dormant seed.

In some areas that were highly visible yet badly degraded, such as the State War Memorial focus site, large grass trees Xanthorrhoea preissii and zamia palms Macrozamia riedlei were planted to enhance the area, many of these were several hundred years old. These large plants provided additional structure as well as photographic opportunities for visitors, and an introduction to our indigenous flora. The grass trees were salvaged from a nearby site, grown on for one year, monitored for root pathogens and then planted. All plants were watered regularly by trickle irrigation for a minimum of 2 years. They were also maintained by the company they were purchased from during the establishment phase.

To make the operation of planting seedlings as efficient as possible, a new plant carrying trailer was purchased. This unit enabled 12,500 seedlings to be transported directly onto the site and was an excellent investment considerably reducing planting costs. Where stabilization matting is laid, holes are cut the day before planting and seedlings are placed on the matting next to the hole. This allows a more efficient operation when using unskilled volunteers to plant. On level ground, to avoid bending over, volunteers used kidney baskets and special planters. Different plant species are used for the top, middle and bottom of the scarp.

Undertaking restoration works on near vertical slopes has never been attempted before in Western Australia and the project has required a range of new skills and work practices to be employed to operate safely in such an environment. Both fixed ropes and fall arrestors were used when working on steep slopes. Once set up by an ‘accredited expert’, both systems are easy to use and allow you to move freely sideways and up and down the slope. Fall arrestors work in a similar way to seat belts. A 30 m cable is attached to an inertia reel inside a metal casing anchored at the crest of the slope. As soon as the worker falls faster than 1.5 m/s, the cable locks off and the fall is arrested.


The majority of on-ground restoration works are now complete. Probably the most challenging task ahead is the control of environmental weeds. This commitment will be ongoing at least until the gaps are filled with indigenous species. With the removal of major weeds, new weeds such as silver grass Vulpia species, fumitory Fumaria species and Centranthus macrosiphon are emerging and filling in the gaps. This has required further research work to find suitable herbicides/techniques to control them without affecting non-target species.

When embarking on any restoration project it is vital to accurately record what has been done and conduct monitoring to measure the success of the project and examine the long term effects as well as securing sufficient funding for long term maintenance.