Neighbourhood-defined approaches for integrating and designing landscape monitoring in Estonia.

摘要

Landscape monitoring is a rapidly developing approach in the field of environmental science and management. In order to develop a sound landscape monitoring programme, key theoretical concepts and study objectives should be clearly stipulated, and the specific objects to be monitored, as well as the criteria for selecting study areas, hierarchical levels, and techniques of data collection and analysis should be identified. This paper describes the development and implementation of the Estonian monitoring programme for agricultural landscapes, conventional approaches for landscape monitoring, and by neighbourhood analysis, assesses how landscape features are covered by different complementary monitoring data and how the current pattern of monitoring networks represents the landscape features. A spatially explicit method of network design for monitoring and sampling strategies combines stratified and multi-scale agricultural landscape monitoring and uses neighbourhood analysis characterised by the nearest neighbourhood index and Ripley's K-function. Data for landscape analysis are obtained from landscape monitoring (three sets) and other complementary environmental monitoring sets, such as biodiversity, forest, soil, and water monitoring (11 sets). It is shown that several monitoring sets follow an approach that aims to achieve national geographical coverage, representing various landscape types. Small sets having less than 50 stations are biased and the networks are not statistically significant. Proportional stratified sampling requires fewer sites for large homogenous inland landscape districts. The concept of agricultural landscape monitoring was tested in pilot areas. The chosen multi-scale object-based methods provide a good overview of the level of human pressure on different categories of agricultural land. Results of the monitoring showed that the species composition and abundance of bio-indicators was, to a great degree, determined by landscape structure. A systematic approach focused on landscape classes helps to integrate the monitoring set as a whole and to achieve a coherent and efficient layout of monitoring sets for Estonia..