MEPS 267:1-8 (2004)  -  doi:10.3354/meps267001

ABSTRACT: A new approach to assess environmental integrity based on PRC analysis was proposed, tested, validated and developed. Environmental health assessment and community studies usually result in complex biological data sets. In order to find ecologically relevant patterns and tendencies from such sets of data it is necessary to reduce all the information to a summarised and simplified form, which might be more easily interpreted by ecologists, politicians, end-users and the population in general. However, several multivariate ordination methods currently used (e.g. redundancy analysis, principal component analysis, or multi-dimensional scaling) produce complex diagrams for the non-ecologist, which do not allow changes in biological communities over time to be easily understood. Here, we propose a recently developed method, principal response curves (PRC) analysis, to overcome these issues. This method has advantages over traditional ordination techniques, or any biotic index, in that it provides a powerful statistical analysis of temporal data series along spatial gradients. The PRC technique can make use of non-disturbed or unpolluted areas as reference sites with which other areas are compared, making it possible to assess changes in species composition between different areas over time. Moreover, individual species responses to stress agents can be inferred from the PRC curves. As well as providing insights into the behaviour of natural ecosystems‹in particular, how ecosystem integrity changes over time‹this new approach can potentially provide a practical tool for monitoring and implementing environmental policy instruments.

KEY WORDS: Environmental integrity · Environmental health · PRC analysis · Community studies