Water Journal : Water Journal April 2012
catchment management technical features 112 APRIL 2012 water Wildfires in the Upper Catchment In February 2009, catastrophic wildfires impacted on a significant area of the upper/mid Goulburn River catchment between Kilmore, Wandong and Alexandra (see Figure 2). The largest of the fires, known as the Kinglake Complex, burnt over a total of 255,000 ha of land (including 155,000 ha in the Goulburn River catchment), destroyed more than 550 homes and resulted in significant loss of life. The fire followed a path across Victoria's Central Highlands, from Wandong, south as far as St Andrews, and east and north through Marysville, Taggerty and Flowerdale towards the upper Goulburn Valley. This fire event has been titled by some as "Australia's worst natural disaster", with the disastrous loss of life, damage to communities, destruction of manmade and natural infrastructure and assets. Over 49% of the fire area was affected by moderate to high soil burn severity. Along with 52% of the area having steep slopes, this put many areas both within and downstream of the fire at increased risk of erosion and runoff (DSE, 2009). While the fire did not directly affect the Goulburn River, its major and minor tributaries were affected. The 2009 wildfires were additional to a series of wildfires in 2003 and 2006 that burnt over much of north-eastern Victoria. Impacts of Wildfires on Catchments Wildfires can have substantial impacts on the characteristics of catchments. Wallbrink et al. (2004) identified potential effects of fires on vegetation and soil cover, physical changes to soil properties after fire and their effect on soil erosion and soil wettability, and the effect of fire on erosion rates. In a study of fire impacts on hydrology and water quality in a wet eucalypt forest environment Lane et al. (2009) found discharge increased by around 70% after fire, and this persisted for at least three years. While there was no apparent change in the runoff processes delivering water to the stream network, suspended and coarse sediment fluxes increased by 8--9 times in the first year post-fire, but relaxed to pre-fire levels by the end of the second year. Phosphorus (P) and nitrogen (N) fluxes increased by approximately 5--6 times, and showed the same recovery rate as sediment, with the majority of both P and N transported in the particulate form. Water quality recovery was a function of the groundcover recovery. Hillslope process experiments revealed the importance of soil water repellency and the spatial arrangement of saturated hydraulic conductivity in pollutant pathway length. These experimental data suggest near-stream areas to be the pollutant source areas. Event water-quality monitoring data indicates the possibility of enormous increases in instantaneous concentrations of water quality parameters. For example, downstream of extensive wildfires in the Mitta Mitta River in 2003, event concentrations of total phosphorous (TP) were up to 5.5mg/L, total nitrogen (TN) up to 65mg/L and suspended solids (SS) up to 45,000mg/L, contrasting with median background levels at the monitoring site of 0.02mg/L, 0.18 mg/L and 3.0mg/L for TP, TN and SS respectively (Ecowise Environmental, 2006). A severe storm (150mm of rain in one hour) over a small sub-catchment (10--15km2) of the nearby Ovens River after wildfires in 2003 resulted in a major fish kill in the river and threatened town and rural water supplies (EPAV, 2003). The sediment 'slug' resulting from this storm event had a dramatic impact on water quality. Following the flash flood, there were very large increases in turbidity in sections of the Ovens River, peaking at 70,000 NTU on 27 February 2003 at Myrtleford. By comparison, turbidity levels in the Ovens River are usually well below 10 NTU. Suspended solid levels, which are typically less than 6mg/L in this river system, peaked at 33,000mg/L; that is, 3.3% solids. In the sediment 'slug' dissolved oxygen levels were very low, dropping to 0.1mg/L. Dissolved oxygen concentrations of less than 1mg/L can lead to fish deaths and significantly impact on other aquatic organisms. These conditions contrasted sharply with water quality downstream of the sediment 'slug', which was more typical of the normal condition of the river. Native fish populations were severely impacted by the sediment 'slug'. A post- event survey in March 2003 found a 98% reduction in native fish abundance in the Buckland River (a tributary of the Ovens River). Drinking water treatment plants struggled to cope with the huge suspended sediment loads, resulting in severe water restriction to reduce demand and trucking of water into towns to augment supplies. Potential Impacts in the Goulburn Catchment Against this backdrop, land and water managers in the Goulburn River catchment had major concerns about the potential water quality impacts of storm events occurring over the area burnt by wildfires in 2009. The fire severity and extent of the burnt area, coupled with the unknown intensity, duration, location and spatial extent of storm events means it is impossible to sensibly plan to manage the effects of a single event. Water quality and associated catchment values could be affected by highly turbid, silt- and ash-laden water. The impact on water quality and natural and human assets depends on a number of factors including the timing, severity and extent of the rainfall event, proximity to assets, e.g. drinking water supply offtake, the rate of flow in the mid-Goulburn River (controlled by releases from Lake Eildon) and diversions at Goulburn Weir. Direct water quality impacts from post-fire rainfall events include increases in suspended solids, nitrogen and phosphorus. Additional water quality impacts could include increases in salinity, decreases in dissolved oxygen and changes in water temperature due to changes in shading, and alterations to instream barriers and stream debris due to incineration of timber or post-fire tree fall. Figure 2. Extent of wildfires in February 2009 (source DSE).
Water Journal May 2012
Water Journal December 2011