Designing forested landscapes to provide multiple services.

Abstract Forest services are benefits generated for society by the existence of certain forest types and their attributes. The particular mix of services, and their amount and quality, depend on the condition of the forest resource. Water and nitrogen processes are determined to a great extent by forest management. Streamwater runoff in areas where water is a scarce resource is significantly affected by tree cover and tree age. Old forests may provide better vistas and more suitable habitat than young forests. Such examples illustrate the overriding importance of the particular forest condition which is created by forest management. Most of the world's forests are utilized by humans and this implies that the dynamics of a forest ecosystem is not so much an ecological, but predominantly a cultural problem. The requirements for forest services are manifold and they are not constant over time. Traditional forest planning is based on principles of constancy and long-term stability. Silvicultural programmes were assumed to remain constant for at least one rotation. In reality, however, periodic reorientation and frequent changes of forest policy are quite common. The history of silviculture is not characterized by constancy, but rather by continuous change in policy. Concrete examples of this vicissitude are changes in the preferred silvicultural systems (clearfelling vs selective harvesting; planting vs natural regeneration), the preferred tree species (beech, spruce, 'exotic' species) and the preferred forest structures (even-aged monocultures; uneven-aged multispecies forest). In view of the difficulty in predicting the direction and rate of change, some of the assumptions that have guided forest planning in the past are re-examined. We first define some of the terminology and show that the dynamic development of a managed forest ecosystem is not only an ecological but predominantly an economic and a cultural problem. We then show that the common practice of standardizing silviculture complicates decision-making and is ineffective in providing multiple services, because it assumes that social, economic and environmental conditions remain constant over time. Nyberg (1998) thus proposed greater emphasis on adaptive management, which involves systematic learning on the basis of the results of past silvicultural activities. Such learning may be slow, however. For this reason, it is advisable to use new paradigms of managing forest ecosystems, together with improved modelling tools, which permit accurate forecasting and systematic evaluation of different management options, based on current information about the forest resource. The 'Multiple Path' theory, which may be considered as a particular form of adaptive forest management, provides a suitable basis for designing forested landscapes. The basic idea is not entirely new and has been implemented in various simplified forms, in North America and Northern Europe. Extensions to cover a desired mix of services have been developed for smaller applications, but the concept has never been recognized as a general basis for designing forests with the aim of delivering multiple services in societies which are committed to the standard of Public Choice. Some principles of forest design are presented and outlined in some detail: how to initialize the landscape and link the levels of the spatial hierarchy; how to balance the mix of services and design their spatial arrangement; and how to integrate varied forms of expertise into forest design. Several of the required modelling tools are briefly explained.