Should the wind disturbance patterns observed in natural forests be mimicked in planted forests in the British uplands?

Recent developments in UK forestry policy require the adoption of managemen t practices that maintain and improve the biodiversity of managed forests. One approach is to use natural disturbance in unmanaged forests as a templa te for setting the scale, frequency and pattern of forest operations in man aged forests. This review considers the relevance of this approach for coni fer plantations in upland Britain. The dynamics of British planted forests are compared with the disturbance d ynamics of analogous natural forests with particular reference to disturban ce by strong winds. Western hemlock-Sitka spruce (Tsuga heterophylla-Picea sitchensis) forests in the Pacific North-west of North America and particul arly South-east Alaska provide the most promising comparison. There are few reports on disturbance in these forests, but the regime includes both gap- phase and stand replacement dynamics due to wind. However, the landscape pr oportion and pattern of resulting structural types are not well defined. The dynamics of planted forests in Britain are dominated by rotational patc h clearfelling which results in regular stand replacement and little possib ility of the stands developing beyond the stage of stem exclusion towards o ld-growth. The pattern and timing of felling is driven by economic and visu al amenity considerations rather than by an attempt to mimic natural distur bance patterns. Moreover, the structural complexity and remnant elements (s uch as deadwood, large trees, vegetation patches) left after large scale di sturbance are rarely found after conventional timber harvesting. The authors conclude that natural wind disturbance regimes have potential a s a reference point for management in British upland forests but at present are not relevant as a model to mimic explicitly This is because the biodiv ersity benefits of adopting a 'natural' approach in planted forests are unc lear compared with management guided by other criteria such as rarity. Furt hermore, the spatial and temporal pattern to be mimicked is not sufficientl y well understood. Improved knowledge could inform decisions on the scale a nd distribution of harvesting across a landscape, and modify silvicultural operations to create and maintain the structures and patterns associated wi th natural disturbance. However, further research is needed to quantify the spatial and temporal characteristics of wind disturbance in upland forests in Britain and in natural forests elsewhere.