Mapping riparian vegetation along rivers: old concepts and new methods

Abstract

Several objections have been made to the approach of a sliced environment introduced by the Remote Sensing (RS) and Geographic Information System (GIS) tools. There is no evidence that each discipline can be reduced to a set of layers of spatial information. For most botanists, the quality and hence the value of a vegetation map rests more heavily on the selected system of classification than on any other feature. This paper assumes that the knowledge of the historical trends in vegetation mapping concepts may provide useful insights for improving the GIS approach of vegetation. In the first part, a summary of the debates in the scientific community is presented. First taxonomists were opposed to physiognomists. Then, with the development of the ecosystem concept and the landscape concept new questions arose in the debate: what should be mapped, vegetation, ecosystems or landscapes? Controversies opposed botanists to geomorphologists. Today, patterns as well as processes have to be mapped. However, little has been done on riparian vegetation. The second part of the paper focuses on two specific requirements for RS of riparian vegetation, namely high spatial resolution and spatially-oriented classification algorithms. Both have been neglected in the past. By 1998, improvements in satellite data should stimulate studies on riparian vegetation. However, aerial photographs will remain the best medium for analysing riparian vegetation in detail. In the third part, the discussion focuses on the use of GIS for riparian vegetation studies. Obviously, a vegetation layer cannot show the vegetation in all its aspects. However, if it is based on a sound scientific method, much of the information which is stored in an implicit form can be exploited for broader application. Mapping vegetation can be considered as a driving element in research, a mean to moderate the subjectivity of conceptual statements and to validate ecological theories. On floodplains the major problem is a lack of useful data and the high cost for obtaining such data. The challenge is to map flood disturbances and the vegetation dynamic.