This Final Report has presented descriptions of the methodology, results, key findings and conclusions of individual tasks undertaken for the SUSDEV 21 Environmental Baseline Survey on Terrestrial Habitat Mapping and Ranking based on Conservation Value. Discussions on the ecological field survey results and their implication on the mapping and ranking of the habitat map were also included in this report. The findings and conclusions of this baseline survey are summarised below:

• A total of 25 habitat mapping categories were defined and used for habitat mapping. An indicative ecological value (high, medium, low and negligible) was assigned to each of these habitats (see TR1 and Section 2 of this report).

• A conservation ranking system was devised to provide an accepted, composite means by which the conservation values of different areas, representing different features, can be ascribed, mapped and compared (see TR1 and Section 3 of this report).

• A preliminary habitat map was produced based on a combination of remote sensing and GIS techniques, using satellite imageries, aerial photos and existing/digitised data. The map was used to help to identify data gaps and formulate ecological field survey strategy (see TR2 and Section 4 of this report).

• A comprehensive review of existing information on the indicative high, medium and low ecological value habitats was undertaken. The aim was to evaluate the adequacy of existing ecological information on each of the "non-negligible" habitat types. This review helped to identify outstanding information gaps which would require field surveys to fill. The information gaps were identified in TR2 and are summarised in Table 10.1a.

• A total of 140 days of survey effort were allocated to the indicatively high, medium and low ecological value habitat types. The strategy for survey effort allocation was proposed in TR2 and actual allocation was presented in this report (Section 5.2). A total of 93, 39 and 8 days were spent, respectively, on the high, medium, and low ecological value habitats.

• Survey sites were selected and proposed for field surveys (TR2) in an attempt to fill the information gaps identified for the habitats. Survey methodology for habitat verification and ecological value assessment devised for each habitat type were proposed in TR2 and finalised in Annex D of this report.

• The information contained in the HKU Biodiversity CD-ROM database has been used for the present baseline survey: firstly so that the surveyors could preview the species information for sites selected for field surveys (Section 5.4) and secondly, the data in the Rare Plants, Sedges and Fung Shui Woods files were used for locating "rare" plant species on the habitat map (Section 7.2).

• Ecological field surveys commenced in August 1999 and were completed in March 2000. A total of 1,051 sub-sites were visited during the 140 ecological field survey days. Wherever possible and applicable, habitat verification and ecological value assessment were conducted during field visits. Data collected from field surveys were analysed and used for editing the preliminary habitat map and refining ecological value ranking (see Sections 6 and 7). Information gaps filled by ecological field surveys were described in Sections 6 and 7 of this report and summarised in Table 10.1a. Information gaps yet to be filled and recommendations for further habitat verification and ecological value assessment are also provided in Table 10.1a.

• The 10-days of survey effort initially assigned to heritage surveys were used to digitise existing heritage information provided by AMO. There were altogether 67 Declared Monuments, 8 Deemed Monuments, 457 Listed (Graded I, II and III) Historic Buildings/Structures and 206 Archaeological Sites digitised into the GIS. A "+" was given to a habitat polygon where a digitised record(s) of any of these heritage features had been located to indicate the presence of a heritage value feature(s) (see Section 8.1).

• Opportunistic notes of any observed recreational and landscape value features were made at sites during the ecological field surveys. The data were assessed against the respective criteria set forth for recreational and landscape value and located on the habitat map (see Sections 8.2 and 8.3). Existing information on recreational/landscape features was reviewed and data which provided grid locations of valued recreational/landscape feature(s) were included for recreational/landscape ranking.

• Habitat verification and ecological value assessment data from field surveys were used to update the mapping and ranking of the habitat map. The updated habitat map comprised a total of 89,542 habitat polygons and a total land cover (above low tide mark) of 111,711.8 ha. Among the 25 habitat categories mapped on the habitat map, Grassland is the most extensive habitat whilst Seagrass Bed occupied the smallest land cover.

While the information contained in the existing habitat mapping system is considered comprehensive and provides adequate baseline information for quantification and evaluation of existing natural habitats, the Government may consider digitisation of additional data onto the habitat map to augment the knowledge database within the system. For example, grid locations of the "biodiversity hot-spots" may be added as a separate GIS layer onto the habitat map once the information is available for data input. Species data contained in the HKU Biodiversity Survey CD-ROM can be applied to the habitat mapping system when a more updated version of the CD-ROM, of which the gaps and inconsistencies within the animal and plant groups' data files have been fixed and species records validated, is available. The locational records of species, eg rare animal species (if identified by HKU specialists), can be overlaid as a separate layer onto the habitat map so that habitats where the animal species have been recorded can be identified. The "rare" species information can also be used for updating the conservation ranking of a habitat if a "rare" species has been recorded within a habitat polygon which has an ecological value lower than high.

The mapping and ranking exercise executed for this SUSDEV 21 baseline survey has produced a comprehensive and robust habitat mapping system which is believed to be a fundamental tool for planning future development in Hong Kong. The habitat map, which forms one of the GIS layers available on the CASET System, allows users to display different map themes and assist in visualisation of the effects of proposed changes (see Annex J for details on the database structure of the habitat map). The habitat mapping system is particularly useful in acting as a tool for quantification and assessment of existing natural habitats. The conservation ranking information provided in the system will facilitate sustainable development planning by highlighting important areas for protection and evaluating the existing baseline conditions against any indicators (developed under the SUSDEV 21 Study) involving the area of land with various levels of conservation status. The SUSDEV 21 Study Team believes that the habitat mapping system will form the basis for future Government policies on land use and natural resource protection which is an important element for sustainable development in Hong Kong.