Publications

3 Environmental research and monitoring (continued)

3.3 Landscape analysis

• 3.3.1 An ecological risk assessment of the major weeds on the Magela Creek floodplain, Kakadu National Park

This programme conducts landscape-scale analysis and monitoring to help differentiate mining related impacts from other causes, which contributes within the broader context to the monitoring of the natural World Heritage values of Kakadu National Park. It responds to recommendations made by the Independent Science Panel in its final report to the World Heritage Committee about mining activity and associated issues in the Alligator Rivers Region, including the possibility of major changes unrelated to mining and the fact that there may also be unforeseen problems arising from mining. There are two parts to the Landscape Analysis of Impacts Program: (1) the development of a conceptual transport pathways model for on-site management within a risk assessment and management framework; and (2) the assessment of natural World Heritage values in the Alligator Rivers Region (such as waterbirds and wetlands, coastal environments, key habitats and species).

Key activities in 2003-04 include:

• toxic and ecological risks of the herbicide tebuthiuron to tropical freshwater species was quantified and accepted for publication in two peer-reviewed international journals;
• presence and significance of endocrine disrupting compounds (EDCs) in recreational water holes in Kakadu National Park, and from stormwater and treated sewage from the Jabiru township, were assessed.
• an ecological risk assessment of three major wetland weeds (para grass, mimosa and salvinia) on the Magela floodplain was completed, and included development of a Decision Support Tool to help managers determine the likely control costs and environmental benefits of different management options;
• a strategic framework to manage feral pest impacts on Kakadu National Park was developed in collaboration with Charles Darwin University (Key Centre for Tropical Wildlife Management), Parks Australia North and Traditional Owners;
• assessment of natural World Heritage values on Kakadu National Park for waterbirds and endemic aquatic invertebrates found in stone country seeps and springs;
• development of seasonal population models for keystone waterbird species to help differentiate landscape-wide impacts of invasive species and climate change from potential mining-related impacts;
• a two-year study on the influence of traditional fire management practices on Boggy Plain, used to enhance natural and cultural wetland values, was concluded. This study will be used as a reference to compare multiple landscape-scale impacts with those on the Magela floodplain.

3.3.1 An ecological risk assessment of the major weeds on the Magela Creek floodplain, Kakadu National Park

In 2000 the Independent Science Panel (ISP) and the International Union for Nature Conservation (IUCN) stated that, whilst impacts from the site-specific Jabiluka mine proposal were most likely very small or negligible, a more comprehensive risk assessment of both the freshwater and terrestrial ecosystem at a landscape-catchment scale was needed. This was because the Alligator Rivers Region is subject to major seasonal or long-term changes unrelated to those that might arise from mining impacts. For example, the impact of invasive weeds is probably the most significant non-mining threat facing the Magela Creek wetland system at the landscape-scale. It was suggested that comprehensive monitoring programmes and accompanying analyses (assessments) are needed to distinguish between impacts from these differing causes and any unforeseen problems arising from mining.

Some weeds have little or no apparent impact - others can be aggressive invaders that dominate ecosystems by reducing available habitat and resources for native plants and animals and negatively impact on the activities of humans. Three such species of concern are the wetland weeds mimosa (Mimosa pigra), salvinia (Salvinia molesta) and para grass (Urochloa mutica). In this context, a risk assessment of these weeds was undertaken for the Magela floodplain. The outcomes of the risk assessment will also be integrated into the conceptual model of pollution pathways for Ranger uranium mine, currently being developed by eriss.

Fortunately, the Magela floodplain remains free of mimosa infestation because Kakadu National Park managers have an active control programme; however, there is a constant threat of invasion from large infestations on the Gunbalanya floodplains just north of Kakadu. Salvinia first appeared in 1983 and has spread throughout the entire Magela floodplain. Nevertheless, its impact is variable due to the highly seasonal nature of its distribution and abundance and the effects of a biological control agent. Salvinia now covers only a small percentage of some billabongs on the floodplain, compared with up to 100% coverage in some areas a decade ago.

Hence, the risk assessment is mainly focused on the ecological impact of para grass, which has been in the region since the 1920s and is now rapidly expanding into many areas on the Magela floodplain (Figure 3.4).

Figure 3.4 Para grass infestation on the Magela floodplain

This may be due in part to the lack of feral buffalo that once grazed on para grass and acted as a form of weed control. The ecological impact of para grass is substantial because it has can form dense monocultures, out-competing native grasses that are essential for the food supply and nesting of waterbirds. Other impacts include:

• altered hydrological regimes;
• altered fire regimes, leading to a possible loss of riparian and monsoon forests;
• impeded movement of larger, aquatic animals;
• reduced availability of, and access to, traditional Aboriginal bush tucker on the floodplains.

An earlier eriss study showed that in the five years between 1991 and 1996, the most heavily infested area of para grass on the Magela floodplain expanded by 290 ha (132 ha to 422 ha). The current risk assessment intends to update the extent of para grass and determine what areas of the floodplain are susceptible to invasion. Fieldwork in 2003 and 2004 has indicated that the extent of para grass on the Magela floodplain has increased considerably since 1996 (Figure 3.5), although the area of coverage has yet to be quantified.

Figure 3.5 Region of the worst para grass infestation on the Magela floodplain (data overlayed on IKONOS™ 2001 image)

QuickBird satellite imagery of the Magela floodplain has recently been acquired and is expected to show the true extent of para grass. This remote sensing technique will help map para grass over different temporal and spatial scales and, in conjunction with a Habitat Suitability Model (HSM), provide a valuable monitoring tool for Park managers.

Researchers from Charles Darwin University are assisting with the development of a Bayesian HSM that will be used to model the relationship between para grass distribution and wetland vegetation. The procedure calculates the relative probability of para grass being present in each vegetation category based on the current distribution. It tests the significance of the results by comparing the actual distribution with a random sample. To date, only the 2003 para grass data have been incorporated into the model. When combined with updated vegetation mapping, also being conducted by eriss, the inclusion of the 2004 data will yield an even more reliable prediction of the invasive capability of para grass on the Magela floodplain.