THE CLASSIFICATION OF NATURAL AND ANTHROPOGENIC VEGETATION IN MADAGASCAR A NEW APPROACH TO VEGETATION CLASSIFICATION AND MAPPING IN MADAGASCAR In Madagascar, the application of the kind of careful, systematic analysis developed by White is critically needed. However, as White (1983) and many others have pointed out, a physiognomic classification for a given area must be developed and tested on the basis of direct field experience (although modern remote sensing techniques can serve as a powerful complement). More importantly, careful chorological analysis is absolutely dependent on a thorough understanding of the taxonomy of the native flora, and also on reliable specimen based distributional data (White 1993). Much work remains to be done before such an approach can be implemented in Madagascar, although the recent progress in botanical inventory work (now mostly being conducted by local collectors) and systematic studies by Malagasy and foreign botanists, is certainly encouraging. We can, however, already formulate a few ideas of what the application of White's basic approach might tell us here in Madagascar. Cornet (1974) developed a bioclimatic map that clearly shows the presence of highly localized climatic conditions in several parts of the country, including in the transition zone around Vohemar located between the humid east coast and the dry north of the island, an area that corresponds precisely to a restricted zone containing a number of locally endemic plant species. Similarly, the transition between the humid east coast forest and the dry south to the west of Fort Dauphin, as shown here at Parcel 3 of the Andohahela Strict Nature Reserve, harbors numerous endemic species including Dypsis decaryi. Cornet's map also shows a specialized local climate around Morafenobe, where a disjunct area of true evergreen forest occurs at Ambohijanahary Special Reserve, with many species that are otherwise restricted to the eastern part of the country. The application of White's approach to Madagascar would also result in treating all evergreen forests as a single broad physiognomic category -- whose distribution, largely coincides with the humid and subhumid stages indicated in blue and green on Cornet's map. Reliable chorological information would then potentially allow us to differentiate floristically distinct zones within these evergreen forests. If the ultimate goal of classification and mapping is to reveal biogeographic patterns in order to recognize and delimit areas that are biologically distinct, then chorology clearly must play a central role. This certainly appears to be the case in Madagascar, where one of the most important applications of a vegetation map is for conservation planning. << BACK | TABLE OF CONTENTS | NEXT >>