From earlier ecological studies it has become apparent that simple univariate or bivariate statistics are often inappropriate, and that multivariate statistical analyses must be applied. Despite several difficulties arising from the application of multivariate methods, community ecology has acquired a mathematical framework, with three consequences: it can develop as an exact science; it can be applied operationally as a computer-assisted science to the solution of environmental problems; and it can exchange information with other disciplines using the language of mathematics. This book comprises the invited lectures, as well as working group reports, on the NATO workshop held in Roscoff (France) to improve the applicability of this new method numerical ecology to specific ecological problems.

Autorentext
Pierre Legendre, geb. 1930, ist Rechtshistoriker mit einem weiten psychoanalytischen, kultur- und religionshistorischen Hintergrund. Er lehrt an der Pariser 'École pratique des Hautes Études'.

Inhalt
I. Invited Lectures.- Scaling techniques.- to ordination techniques.- Some multidimensional scaling and related procedures devised at Bell Laboratories, with ecological applications.- The duality diagram: a means for better practical applications.- Nonlinear multivariate analysis with optimal scaling.- Joint ordination of species and sites: the unfolding technique.- Clustering under a priori models.- Some non-standard clustering algorithms.- Constrained clustering.- Species associations and conditional clustering.- Fractal Theory.- Applications of fractal theory to ecology.- Path analysis for mixed variables.- Path analysis with optimal scaling.- Spatial analysis.- Spatial point pattern analysis in ecology.- Applications of spatial autocorrelation in ecology.- II. Working Group Reports.- Numerical ecology: developments for microbial ecology.- Numerical ecology: developments for studying the benthos.- Data analysis in pelagic community studies.- Numerical ecology: developments for biological oceanography and limnology.- Numerical methods in terrestrial plant ecology.- Novel statistical analyses in terrestrial animal ecology: dirty data and clean questions.- List of participants.