Résumé
Ecological indicators are increasingly used to examine the evolution of natural ecosystems and the impacts of human activities. Assessing their trends to develop comparative analyses is essential. We introduce the analysis of convergence, a novel approach to evaluate the dynamic and trends of ecological indicators and predict their behavior in the long-term. Specifically, we use a non-parametric estimation of Gaussian kernel density functions and transition probability matrix integrated in the R software. We validate the performance of our methodology through a practical application to three different ecological indicators to study whether Mediterranean countries converge towards similar fisheries practices. We focus on how distributions evolve over time for the Marine Trophic Index, the Fishing in Balance Index and the Expansion Factor during 1950-2010. Results show that Mediterranean countries persist in their fishery behaviors throughout the time series, although a tendency towards similar negative effects on the ecosystem is apparent in the long-term. This methodology can be easily reproduced with different indicators and/or ecosystems in order to analyze ecosystem dynamics.