Open Access Repository

Bayesian Decision Networks for Management of High Conservation Assets (National Water Initiative – Australian Government Water Fund. Report 6/6 Report to the Conservation of Freshwater Ecosystem Values Project, Water Resources Division, Department of Primary Industries and Water)

Downloads

Downloads per month over past year

Davies, PE (2007) Bayesian Decision Networks for Management of High Conservation Assets (National Water Initiative – Australian Government Water Fund. Report 6/6 Report to the Conservation of Freshwater Ecosystem Values Project, Water Resources Division, Department of Primary Industries and Water). Technical Report. DPIWE, Hobart and Freshwater Systems, Hobart.

[img]
Preview
PDF
AGWF_CFEV_BDN_r...pdf | Download (1MB)
Available under University of Tasmania Standard License.

| Preview

Abstract

The Conservation of Freshwater Ecosystem Values (CFEV) Validation project,
funded under the Australian Government Water Fund (AGWF), assessed the validity
of a number of attributes of biophysical character and condition in the CFEV database
for selected river and wetland assets of high conservation value in five catchments.
High Conservation Value assets which are affected by existing Water Management
Planning had been selected for field inspection, and observations were compared,
where relevant, to values attributed to them in the CFEV database. As part of this
exercise, detailed descriptions of each asset were provided, and accompanied by a
description of the major ‘drivers’ of condition evident at each location, and the main
management issues and priorities for the conservation of the asset’s values. These data
were summarised in tabular form and were also accompanied by a summary of the
priorities for Water Management Planning.
Thus the project conducted an ‘on-ground’ validation of asset characteristics, and
reported on the biophysical classes and condition of assets within each catchment and
described their management needs.
A final aspect of this project was to initiate the development of decision support for
management of these assets by developing preliminary Bayesian Decision Networks
(BDN’s) informed by the learnings from the field surveys and from the ‘expert’
project team (Dr Peter Davies, Dr Lois Koehnken and Dr Philip Barker).
Bayesian networks are able to integrate different sources of evidence, represent
uncertainties in knowledge and inherently variable environments and explicitly link
ecological outcomes with management activities and system context and changes. The
primary aim of this project component was to develop Bayesian network models for
predicting the ecological condition of stream and wetland assets, using biological
condition as the primary indicator of asset condition or ‘health’.
The networks developed here identify the dominant linkages between management
actions and the physical and biological components of the ecosystem assets. Two
generic networks have been produced that can be adapted to form the basis of a guide
to making decisions about asset management, as well as to examine changes in
biophysical attributes in response to controllable system changes, such as
management actions.

Item Type: Report (Technical Report)
Publisher: DPIWE, Hobart and Freshwater Systems
Date Deposited: 13 Feb 2008 00:30
Last Modified: 18 Nov 2014 03:29
Item Statistics: View statistics for this item

Actions (login required)

Item Control Page Item Control Page
TOP