ABSTRACT
First of all, it should be stressed that the resolution of what was done was coarse; the results are not comparable in their accuracy to those obtained by more conventional methods of aerial photographic interpretation, vegetation mapping, soil survey and land classification. There are two reasons why this is not considered disturbing. Firstly, the area of study was large, (around 45 000 sq. km) and the manpower necessary to conduct the sort of surveys mentioned was not available (although vegetation mapping and soil survey are now under way in Tsavo) as had been the fortunate case in the Serengeti (e.g. Gerresheim, 1974; de Wit, 1973; Herlocker, 1974). In addition, the aerial photo-cover of Tsavo, although complete, was found (Norton-Griffiths, 1973) to be of inadequate quality to fulfil the requirements of this type of survey. Secondly, the sort of information that was required, namely that for interpreting animal distribution, could for the most part be collected as a part of animal counting exercises; certain information had, of necessity, to be collected as a separate exercise, in particular climatic data, while other aspects of the system, such as the soils and topography, could always be added to the store of information as they became available.
