Wider Catchment Impacts



By Frank Batini (Forester) and Keith Barrett (Hydrologist)

Reproduced, with permission, from The Forester Volume 50 Number 4, Dec 2007. This is an update on the effects of the wildfire described by Dr C Terry at the Eaton Seminar 2005.


In January 2005 a bushfire in the Perth hills area burnt approximately 27,700 hectares. An arsonist lit several fires to the south of the Mundaring Weir reservoir, in steep country and under windy conditions. Despite early suppression efforts by CALM (now DEC) staff, several of these fires coalesced into a wind-driven inferno that headed west towards Pickering Brook, Roleystone and near-by suburbs.

Perth residents awoke to a dense smoke haze that lasted for several days.

Suppression efforts continued for over a week and involved over 700 fire-fighters (from DEC, Forest Products Commission, FESA and volunteer Brigades) as well as aerial suppression using both helicopters and aircraft, A total area of about 27,700 ha of jarrah and wandoo forest and P. radiata plantation; on three water-supply catchments (Mundaring, Canning and Victoria) was burnt. Fortunately no lives were lost and no houses were burnt, though damage to sheds and orchards occurred and a number of firefighters were injured.  The 700 hectares of burnt pine forest managed by the FPC required salvage.

DEC fire manager Rick Sneeuwjagt said later that strategic prescribed burns carried out by CALM 1 -3 years earlier had allowed his staff to control fires that were critically endangering suburbs such as Roleystone.

Experience in Sydney and Canberra catchments over the last few years indicate the potential for this bushfire to impact on the quality of Perth’s water supplies. Potential impacts include:

  • short-term increases in surface runoff and yield due to reduced evapotranspiration;
  • increased turbidity due to loss of vegetative cover resulting in erosion of catchment soil and streambeds;
  • increased colour of water;
  • increased organic material and potential decrease in dissolved oxygen;
  • release of nutrients and consequent eutrophication; and
  • Altered surface soil properties as a result of the fires causing the soil to repel water.

In the Mundaring Catchment, areas adjacent to the reservoir and a large area of steep land in the Darkin River sub- catchment were burnt by very high intensity fire. As Mundaring is the main water source to the Goldfield and Agricultural Regions, the resultant fire impacts were of critical concern. Well-planned remedial action was required and needed to be in place before the coming winter reins.


A Post Wildfire Action Plan was developed by the Water Corporation staff with input from experienced foresters, hydrologists and engineers The objective was to mitigate adverse impacts of the bushfire on water quality.

Major actions stemming from the Action Plan included :

  • field survey of the fire area to identify and rank potential problems;
  • development and implementation of a water quality monitoring program;
  • implementation of erosion and sediment control measures;
  • a review of reservoir operating strategies and preparation of a contingency plan in the event of serious water quality problems; and
  • liaising with land managers to direct remediation actions.


There were no significant impacts to water quality at the reservoir wall Heavy rainfall from late March 2005 resulted in sediment and ash moving downstream within the catchment and a plume developed within the upper part of the reservoir. In this long, narrow reservoir for each event the plume was assimilated well before the take-off point. Erosion control structures within the catchment were able to trap some sediment load reducing the total load reaching the reservoir. Response measures were proposed should water quality reach levels of concern in Mundaring Reservoir. Fortunately, this did not occur.


Baseline rainfall, water quality and yield data were available for two adjacent sub-catchments, Little Darkin (severely affected by the wildfire) and Pickering Brook (prescribed burning some 12-18 months previously, thus allowing the wildfire to be controlled). The two v-notch gauging stations were re-opened by the Water Corporation after the fire and pumping samplers installed.

Three months after the fire, fifteen plots were established, nine in the burnt areas (stratified by burn severity as classified by DEC) and six in unburnt areas as controls. At each site data were collected on damage and recovery of the overstorey trees, on understorey density, structure, dominant species, on litter, slope and soil. These plots were re-measured 14 months later.

Data on invertebrate terrestrial fauna were collected by Ms Chantelle Jackson from the University of WA at various sites in both catchments. Data on aquatic biodiversity were collected from several streams during the 2005 winter by staff from DEC’S Science Division.


Soils– About 12% of the soil surface in burnt areas was visibly “baked” and altered by the fire, thus allowing for increased water-shedding, with 80% of burnt plots showing evidence of surface soil erosion. There was no evidence of surface soil movement in unburnt areas.

Water quality– Following rainfall events in March and April when about 40 mm and then 60 mm fell on the catchments, very large quantities of soil, silt and ash were deposited within stream-beds and pools. Subsequently, in the stilling pond at the Little Darkin v-notch weir, an estimated total of 350 cubic metres of soil and organic matter were removed on three separate occasions during winter 2005 to keep the v-notch operational. Large amounts of fine silt and ash also flowed over the v-notch and into the Mundaring Weir. By winter 2006, which was very dry, the soils were stabilised by vegetation and much less silt was deposited in the streambed.

Water Yield– In the first winter after the fire, the burnt catchment yielded a flow that was 2.2 times that which would be expected based on historical comparisons. This catchment reacted swiftly to rainfall events in the early part of winter when the ground was bare and the soils were water shedding. In the dry winter of 2006, flow in the burnt catchment was only marginally higher. Monitoring is expected to continue for 5-8 years, since flows in the wildfire-affected areas are expected to decrease markedly as the tree crownsrecover and the dense regeneration takes hold.

Aquatic biodiversity– during the 2005 winter, sampling by DEC showed that the aquatic biodiversity in the Little Darkin stream was significantly impaired.

Terrestrial invertebrates– data collected in winter 2006 show that invertebrates within the wildfire areas were less diverse and contained only a sub-set of the species found in the area that had been previously prescribed burnt.

Birds and mammals– In contrast to areas that had been prescribed burnt, during the first winter there were no frog- calls or birdsong and few mammal tracks were observed in the areas burnt by high intensity fires.

Tree deaths– it is estimated from plot data that between 1.5 and 2.3 million trees greater than 10 cms in size were killed in the wildfire. In areas that were fully defoliated about one-third of the trees were killed. Many of these were large, habitat trees.

Recovery-Some eighteen months after the fire, the understorey density had recovered to levels observed in control plots. Density of tree crowns had recovered well in areas that were scorched. Defoliated areas showed epicormic growth but tree canopies are still below levels in the controls.


The very large wildfire in the hills catchments in January 2005 had potential for severe consequences on water quality in domestic water supply reservoirs. In subsequent years this wildfire is expected to also impact negatively on water yield.

Streams and pools were filled with silt thus significantly impairing aquatic fauna. As these are low-energy streams, the silt will remain for years. Many habitat trees were killed and most other trees were damaged. The terrestrial ecosystem after the wildfire was more simplistic and less diverse. After 18 months, the forest, understorey and terrestrial ecosystem were beginning to recover, but full recovery may take some decades.


Department of Environment and Conservation (2006) Report

on Key Performance Indicator No 20.

Jackson, Chantelle (2006) The effects of an intense wildfire

on terrestrial invertebrates within the riparian zone of streams

in the Northern Jarrah forest, south-west Western Australia.

(Submitted in partial fulfilment of the degree of Bachelor of

Science (Honours) at the University of Western Australia).


DEC’S Rick Sneeuwjagt and Dr Li Shiu, Hydrosmart

hydrographers, Water Corporation’s catchment rangers, Dr C

Terry and Ms A Reed.

Mr Batini and Mr Barrett are consultants to the Water

Corporation. Email: fbatini@bigpond.net.au