New high nature value map of Estonian agricultural land: Application of an expert system to integrate biodiversity, landscape and land use management indicators

This paper describes the national databases and methods that have been used to construct a new expert system-based tool to map High Nature Value (HNV) agricultural land in Estonia. Twenty appropriate indicators from four thematic groups, i.e., land use management, nature conservation, landscape diversity and inherent natural quality, were selected on the basis of use in the literature, the requirement of consistent national datasets and statistical analysis. Each variable was divided into four appropriate classes to produce indicator values according to expert judgement. These classes were added together for each 1 km square to provide a single score to develop an expert system to define HNV. A statistical analysis showed there are few moderate correlations between the individual indicators, suggesting that their selection was sound. The HNV scores for all 1 km squares with agricultural land in Estonia exhibit a normal distribution. The top and bottom 10% of the 1 km squares (3707 and 3501 squares, respectively) were identified in order to investigate the structure of the data further. The former is termed Exceptionally High HNV (EHNV) and the latter Relatively Low HNV (RLNV), while the intermediate class is termed Median HNV (MHNV). The distribution of these groups in Estonia shows readily identifiable patterns linked to known biodiversity “hot spots”. A cluster analysis within the EHNV squares produced three classes whose distribution and composition are described to demonstrate the detailed character of Estonian agricultural land with high biodiversity. The methodology is adaptable to other countries that have extensive relevant data bases. The map will be valuable for agricultural policy makers to identify zones with high biodiversity where agri-environment schemes can be targeted. Other stakeholders could also direct development away from EHNV 1 km squares.