Résumé : The relation between landscape patterns and ecological processes forms a central hypothesis of landscape ecology. Three types of pattern analysis to assess anthropogenic impacts on landscape ecosystems and biodiversity are presented in this chapter. Firstly, the results of an analysis of Acanthaceae data in Central Africa are presented and compared with phytogeographic theories. Phytogeography data reflect the spatial variability of plant diversity, and constitute therefore a major tool in conservation policy development. We investigated if it was possible to proxy the phytogeographic classifications by the spatial distribution of Acanthaceae only. When combined with a classic landscape pattern analysis, this type of study could provide complementary information for the definition of conservation priorities. Secondly, we present an analysis of periodic vegetations in the Sudan. It can be accepted that through an understanding of the underlying mechanisms of the formation of this unique pattern geometry, the knowledge with regard to the functioning and vulnerability of these ecosystems can be deepened. Using high-resolution remote sensing imagery and digital elevation models, the relation between pattern symmetry and slope gradient was explored. In particular, slope gradients that could condition the transition between spotted and tiger bush pattern types were focused. The influence of other sources of anisotropy was also considered. Finally, a complementary approach to the calculation of landscape metrics to analyse landscape pattern is described, using the spatial processes themselves causing landscape transformation. Landscape ecologists agree that there appears to be a limited number of common spatial configurations that can result from land transformation processes. Ten processes of landscape transformation are considered: aggregation, attrition, creation, deformation, dissection, enlargement, fragmentation, perforation, shift, and shrinkage. A decision tree is presented that enables definition of the transformation process involved using patch-based data. This technique can help landscape managers to refine their description of landscape dynamics and will assist them in identifying the drivers of landscape transformation.