The British Government's multi-purpose forest policy 'seeks to ensure that forests and woodlands are managed and developed so that their social, environmental and economic benefits are realised for the community at large' (Forestry Commission, 1993). In addition, the government has to meet international commitments resulting from the ' Statement of forest principles' agreed at the 'Earth Summit' at Rio de Janeiro in 1992, and the 1994 Helsinki European Ministerial Conference at which agreement was reached to manage European forests in a sustainable 'way and at a rate that maintains their biodiversity (and) productivity...' (Sustainable Forestry: The UK programme,. 1994). The Forestry Commission's Biodiversity Research Programme is providing baseline information about forest biodiversity, and the way in which forestry practice affects biodiversity. Forest managers need to consider an increasing volume of information in order to meet multiple forestry objectives. Decision support (DS) systems can help busy managers select and assemble vital information.
This paper discusses the potential of knowledge-based systems to provide help for the forest manager with site classification, species choice, timber yield and recognition of valuable forest habitat, through Ecological Site Classification (ESC), that will form a core element of sustainable forest management. Input to the prototype DS system will vary from basic locational information (grid reference and altitude) to more detailed site data (soil type, ground flora). The DS system will offer a range of outputs, e.g. site suitability criteria for tree species and for native woodland communities. In addition, the system will provide a report with a step-by-step explanation of the decision process.
The prototype ESC DS system will be developed with the aid of the software Knowledge Pro for Windows (KPWin). The KPWin environment offers essential expert system programming tools such as backward and forward chaining, combined object oriented programming (OOP) as well as list process programming. In addition, dynamic data exchange (DDE), and multi- media support with hyper-text and hyper-regions, will enable the DS system to have a user friendly graphics interface and allow a large amount of reference material to be accessed, as and when required, by novice users. KPWin also supports modular programme construction, allowing new features to be introduced, and linkages with other applications to be made.
Future development objectives aim to make ESC DS a core module of a decision support system offering additional ecological applications on spatial information by means of a linking interface with a commercial Geographical Information System (GIS). Using a loose coupling approach, those queries defined by the user involving any kind of spatial analysis, such as overlay operations and spatial interpolations, will be treated as goals to be satisfied by the system. Such a system could offer information on rare species, the location and extent of particular habitats and on the management of deadwood, and it could perform cost-benefit analyses of forestry planning options.