RRRD027: Getting ground cover right: thresholds and baselines for a healthier reef

Terry S. Beutel1, Dan Tindall2, Robert Denham2, Rebecca Trevithick2, Peter Scarth2,3, Brett Abbott4 and Chris Holloway1

1 Department of Agriculture, Fisheries and Forestry, Queensland.
2 Department of Science, Information Technology, Information and the Arts, Queensland.
3University of Queensland, Centre for Spatial Environmental Research.
4 Commonwealth Scientific and Industrial Research Organisation.

pdfDownload the RRRD027 Research Outcomes Report5.4 MB

Executive Summary

There is a clear and ongoing need in the Great Barrier Reef (GBR) catchments to quantify changes in the catchment landscape, and to understand how much these changes are driven by climate, management and other factors. This information is fundamental to both risk mitigation on the reef and to ensuring sustainable and productive use of the catchment landscape. Getting ground cover right directly addressed these needs in the Burdekin and Fitzroy natural resource management (NRM) regions of the GBR catchment, developing and distributing landscape monitoring resources for reef catchment stakeholders. These resources included:

  • remotely sensed, catchment wide ground cover data;
  • spatially validated grazing land condition mapping; and,
  • customised monitoring support tools for reef stakeholders.

 Project partners have processed over 60,000 Landsat images, converted them to fractional cover imagery and generated a time series of seasonal (3 monthly) cover images for the whole state. The fractional ground cover algorithm developed here estimates ground cover more robustly than its predecessor, and so allows better separation of over storey and ground cover. In addition, we have developed an approach to further separate over storey and ground cover, and developed two tools for distinguishing climatic and management impacts on ground cover levels.

The project also produced a new generation of spatially validated, multi decadal mapping of grazing land condition for the Burdekin and Fitzroy regions. The mapping was validated on over 1700 ground sites, and identified a number of spatial and temporal patterns in grazing land condition across the two regions. The new generation of mapping has been used in local Research, Development and Extension (RD&E) projects. In addition, our validation site data is available for future RD&E looking into the region's landscape health, as exemplified by our own investigations of sustainable ground cover thresholds, of which the validation data were a key component.

The project conducted two analyses of potentially sustainable ground cover thresholds for the two regions. One analysis investigated the relationship between catchment ground cover levels and risk of surplus catchment sediment runoff, but did not produce an informative outcome. A second analysis modeled relationships between relative levels of ground cover in dry times, land condition and long term rainfall on our validation sites. It has generated pixel and sub-catchment scale estimates of levels of cover required to maintain land condition for the two regions, and looks likely to be a help improve the setting of regionally-specific ground cover targets in GBR catchments.

The above products have been packaged and distributed via multiple channels. Most notably, we have packaged the seasonal cover time series in the VegMachine® software and trained 32 RD&E staff to use that software. This has resulted in assessments of landscape state and change on over 300 grazing properties to help direct NRM funding and extension work, and to directly provide land managers with landscape monitoring data.

A key point to note about the project has been its impact on other RD&E work in the GBR catchment. Project outputs have been delivered to numerous end users, and are shaping a large part of ongoing RD&E in the region. Direct impacts include:

  • provision of cover data for catchment modelling;
  • capacity to report on ground cover targets in reef catchments;
  • more informed evaluation of NRM funding applications;
  • provision of information for Water Quality Improvement planning;
  • identification of extension target areas;
  • supply of land managers with historical land monitoring data;
  • more informed economic assessment of land rehabilitation options;
  • identification of appropriate RD&E sites; and,
  • generation of fire history mapping for the state.

This is a sizeable and still growing list of activities that have benefitted from the work funded under this project, and represents a very significant return on that investment.

We recommend three core future RD&E directions from this work. These are discussed in greater detail later in the report, but in summary are as follows:

  1. Secure the Queensland Ground Cover Monitoring Program. The Queensland Ground Cover Monitoring Program's ground cover dataset is foundational to RD&E in the GBR catchment and beyond. At this point in time, the funding future of the program is unclear. We see a clear priority in securing the ground cover program and its acquisition and processing capacity.
  2. Move VegMachine online. The VegMachine PC software has proven its capacity to open up the ground cover data for a wide range of users. With current technologies it is eminently feasible to deliver VegMachine through an online platform. This would circumvent the limitations of the current delivery model, and exponentially expand public access to ground cover data.
  3. Keep leveraging the ground cover data. The ground cover data and its associated products are the most detailed dataset we have of physical change in the GBR catchments. Our work around land condition mapping is just one of many examples (see above) of how these data can inform our understanding of the landscape and our roles and options within it. Future RD&E should look for novel ways to leverage this dataset – it has produced a great deal of value already, but we think there is a great deal more value to be realised.