At present time a strong demand of spatial data exists in the wide range of ecologically oriented disciplines. This demand cant be satisfied only by traditional methods of monitoring and could be complemented by data and techniques of Remote Sensing (RS). Nevertheless common classification techniques currently dont integrate additional ecologically oriented data and results achieved just according to spectral properties of various land-covers are having especially in case of natural landscape (which is full of mixed classes and continuously transitioning land-covers with similar spectral characteristics) low ecological quality and consequently users acceptance. This paper is offering partial solution of above stated problem employing Knowledge Base classification approach. Specific Knowledge Base (KB) for area of interest was built using layers of fuzzy classification and expert knowledge about environmental claims of various plant species. Achieved results have been compared with those coming from common classification methods via confusion matrices and Kappa statistics. KB allows improvements of overall accuracy about 5%. Results of KB classification of one-date imagery are comparable with MLH classification using multi-temporal data.