Representing Indigenous wetland ecological knowledge in a Bayesian Belief Network
Peter Christophersen, Sandra McGregor, Barbie McKaige · 2009
Abstract: It is widely appreciated that Indigenous Australians hold a wealth of ecological knowledge that could be beneficially applied to contemporary land management. However, this has rarely happened, and unfortunately a large amount of Indigenous knowledge is being lost as elders pass away. The Bayesian Belief Network (BBN) approach is ideal for recording traditional ecological knowledge and applying it to land management as it can use qualitative information in the form of expert opinion using local terminology. Once a model is developed, the BBN approach also provides an intuitive means of exploring system dynamics, therefore offering an effective educational tool for Indigenous and non-Indigenous people alike. The collaborative process of model development also fosters new relationships and a better understanding of the ecosystem by all parties. This project was designed to examine the how Indigenous land managers recognise and manage for healthy wetlands. This was achieved by working with a family of Aboriginal land managers, a number of whom are traditional owners in Kakadu National Park. A BBN was developed to formalise the integration of western and indigenous knowledge systems, and to develop a visually appealing, interactive, educational experience for a diverse audience, from Aboriginal land managers to tourists and park management. A web-based, graphical presentation was then developed to clearly present BBNs and display additional underlying knowledge. This paper presents the process and unique challenges in developing the BBN with Indigenous Australians who have not been previously exposed to western modelling approaches.