Kakadu National Park Landscape Symposia Series 2007–2009
|
|  Yüklə 5.42 Mb.
|
6.5 Weed risk assessment modellingMulti-temporal information derived from remotely sensed data can be used to predict invasion rates of weeds in different habitats. A specific example from KNP is shown in Figure 3, in which the rate of increase in extent of para grass on the Magela floodplain was assessed to be 14% per annum or to be doubling in extent every 5 years. Distribution maps and other environmental information relating to habitat preferences of weeds (eg site specific water depth on wetlands) can be also integrated into spatially explicit models developed within Geographic information systems (Ferdinands et al 2001, Ferdinands et al 2005, Ferdinands 2007). 
Figure 3 Estimate of increase in area for para grass on a selected region of the Magela Creek floodplain (as shown in Figure 2b) (from Bayliss et al 2006) 6.6 Conclusions and recommendationsManagers are better able to combat the spread of weeds through an understanding of the spatial and temporal context of weed invasions. VHR remotely sensed imagery can address the information gaps associated with managing weeds in remote and inaccessible landscapes by providing detailed and accurate maps of specific wetland weed species and native vegetation. In particular, if VHR imagery were to be applied for routine monitoring in high-value conservation areas, there would be a improved capacity to detect small ‘satellite’ weed colonies in a timely and cost-effective manner. Remote sensing provides a tool to recognise, understand, and manage change in remote environments, by delivering reliable, defensible, and measurable criteria for mapping weeds and the condition of native habitats. Satellite RS is the most cost-effective source of information for acquiring continuous spatial coverage of vegetation condition over large areas. Such information also contributes to a broader understanding of ecological function of natural environments such as wetlands (Johnston & Barson 1993, Ozesmi & Bauer 2002, Baker et al 2006). RS, coupled with GIS based models, are a basis also for incorporating spatial and temporal knowledge of the landscape into risk assessment decision support tools for managers (Leuven & Poudevigne 2002). While the utility of VHR imagery for mapping para grass has been demonstrated, utility for other target weed species of the region has not been fully investigated. However it is likely that any weed species which forms dense monospecific patches in the landscape, is a candidate for mapping using VHR imagery, provided there are sufficient visual differences (in spectral reflectance, texture, and shape) to distinguish colonies from surrounding features. In this context grassy weeds on wetlands (Olive Hymenachne) and terrestrial woodland (Pennisetum spp) are good candidates for mapping. Management priorities within Kakadu are focused on conserving the unique natural and cultural heritage values (which are tightly coupled with respect to wetlands within the region). In this context maps of weeds and native vegetation (representing habitat management units) can be used as monitoring endpoints for risk assessment and decision support for managers. Such an approach is complementary to the KNP Plan of Management (2007–2014) and will facilitate adopting a ‘habitat-unit’ approach to managing natural resources within the Park at a landscape scale (Director of National Parks 2007). Cost-effective delivery of RS products requires developing procedures, standards, and agreed measurement endpoints under a coordinated framework. In this respect there is a need to refine protocols and allocate resources to monitoring and assessment of change in variable wetland ecosystems (Shanmugam et al 2006). Adaptive weed control operations will also benefit from such a monitoring program, where remote sensing mapping and validation is integrated with spatial knowledge from weed control operations and surveys.
ReferencesUSEPA 1998. Guidelines for Ecological Risk Assessment. EPA/630/R-95/002F. Risk Assessment Forum. US EPA (US Environmental Protection Agency), Washington, DC, USA. Australia, Director of National Parks & Kakadu National Park Board of Management 2007. Kakadu National Park: management plan 2007–2014. Canberra. Baker C, Lawrence R, Montagne C & Pattern D 2006. Mapping wetland and riparian areas using Landsat ETM+ imagery and decision-tree-based models. Wetlands 26, 465–474. Bayliss P, van Dam R & Humphrey C 2006. Ecological risk assessment of Magela floodplain, Kakadu National Park: Comparing point source risks from Ranger uranium mine to diffuse landscape-scale risks. Proceedings of the Society for Risk Analysis (SRA) Conference (Australian and New Zealand Chapters). University of Melbourne: Australian Centre of Excellence in Risk Assessment. Boyden J, Bartolo R, Bayliss P, Christophersen P, Lawson V, McGregor S & Kennett R 2008. Initial assessment of high-resolution remote sensing to map and monitor change in wetland vegetation on Boggy Plains, World Heritage Area, Kakadu National Park, Australia. In 14ARSPC: Proceedings of the 14th Australasian Remote Sensing and Photogrammetry Conference, Darwin NT, 30 September – 2 October 2008. USB2.0 Boyden J, Walden D, Bartolo R & Bayliss P 2007. Utility of VHR remote sensing data for landscape scale assessment of the environmental weed Para grass [Urochloa mutica, (FORSSK), Nguyen] on a tropical floodplain. 28th Asian conference on remote sensing. Kuala Lumpur, 12–16 November 2007. Catt P & Thirarongnarong K 1992. An Evaluation of remote sensing techniques for the detection, mapping, and monitoring of invasive plant species in coastal wetlands: A case study of Para grass (Brachiaria mutica). 6th Australasian Remote Sensing Conference. Wellington, New Zealand. Everitt JH, Fletcher RS, Elder HS & Yang C 2007. Mapping giant salvinia with satellite imagery and image analysis. Environmental Monitoring and Assessment 139, 35–40. Everitt JH, Yang C, Fletcher RS, Davis MR & Drawe DL 2004. Using aerial colour-infrared photography and QuickBird satellite imagery for mapping wetland vegetation Geocarto International 19, 15–22. Ferdinands K 2007. Assessing the threat posed by an invasive African grass Urochloa mutica (Forssk) Nguyen to biodiversity conservation in the Mary River wetlands, Northern Territory, Charles Darwin University, Darwin. Ferdinands K, Beggs K & Whitehead P 2005. Biodiversity of invasive grass species: multiple-use or monoculture. Wildlife Research 32, 447–457. Ferdinands K, Davenport C & Whitehead P 2001. The use of GIS-based predictive models to examine the potential impact of Para grass Urochloa mutica (Forssk) on landscape structure in the Mary River floodplains, NT. North Australian Remote Sensing & GIS Conference. Darwin. Harvey KR & Hill GJE 2001. Vegetation mapping of a tropical freshwater swamp in the Northern Territory, Australia: a comparison of aerial photography, Landsat TM and SPOT satellite imagery. International Journal of Remote Sensing 22, 2911–2925. Johansen K, Coops NC, Gergel SE & Stange Y 2007. Application of high spatial resolution satellite imagery for riparian and forest ecosystem classification. Remote Sensing of Environment 110, 29–44. Johnston RM & Barson MM 1993. Remote sensing of Australian wetlands: An evaluation of Landsat TM data for inventory and classification. Australian Journal of Marine and Freshwater Research 44, 235–252. Leuven R & Poudevigne I 2002. Riverine landscape dynamics and ecological risk assessment. Freshwater Biology 47, 845–865. Menges CH, Ahmad W, Khwaja ZH & Hill GJE 1996. Merging minimum-distance-to-mean and maximum likelihood algorithms for classifying Mimosa pigra in northern Australia. Asian-Pacific Remote Sensing & GIS Journal 9, 51–59. Ozesmi SL & Bauer ME 2002. Satellite remote sensing of wetlands. Wetlands Ecology and Management 10, 381–402. Phinn SR, Stow DA & Van Mouwerik D 1999. Remotely sensed estimates of vegetation structural characteristics in restored wetlands, southern California. Photogrammetric Engineering & Remote Sensing 65, 485–493. Phinn SRSDAZJB 1996. Monitoring wetland habitat restoration in Southern California using airborne multispectral video data. Restoration Ecology 4, 412. Shanmugam P, Ahn Y & Sanjeevi S 2006. A comparison of the classification of wetland characteristics by linear spectral mixture modelling and traditional hard classifiers on multispectral remotely sensed imagery in southern India. Ecological Modelling 194, 379–394. Tuxen KA, Schile LM, Kelly M & Siegel SW 2007. Vegetation colonisation in a restoring tidal marsh: A remote sensing approach. Restoration Ecology 16 (2), 313–323.
|