ABSTRACT
Classification is an essential prerequisite for effective landscape evaluation, planning and management. In examining the representation of individual threatened landscapes in a protected areas network, for example, it is necessary to ensure that similar landscapes are compared with each other, otherwise the range of variability is so great that comparisons are invalid, as described by Watkins & Bunce (1996). Thus it is not possible to compare a Mediterranean landscape in Crete with a boreal landscape in northern Norway as they are inherently different. Such comparisons commonly have policy implications. Some of the landscape classes that are identified may have no representation of designated sites, either because of national policies, or because they have been overlooked. In Britain, a good example of this problem was provided by the setting up of National Scenic Areas in Scotland. The selection was made by a process of expert judgement. However, it completely omitted the Flow Country of the extreme north. These flat boglands, though visually uninspiring to many people, were subsequently identified as being unique in landscape terms as well as for their biodiversity. They were ignored because of their lack of striking mountain features.
