Natural disturbance dynamics in an old-growth forest: from tree to landscape

Forest structure is the result of past disturbances, regeneration establishment, competition, and mortality dynamics. Natural disturbances can create single dead tree to larger canopy gaps creating both small and large scale patterns and processes. The analysis of forest spatial pattern can provide information on these dynamics shaping forest ecosystems. Knowledge about primeval forest dynamics is relevant for their conservation and is an important reference for delineating sustainable forest management. The main aim of this research was to assess forest structure and understand spatial patterns and related disturbance processes in a mixed silver fir-Norway spruce-beech old-growth forest. The study site is the forest reserve of Lom, an old-growth forest located in Bosnia and Erzegovina. We analysed forest structure at different scales and with different approaches : a) landscape, b) forest, and c) single tree. A Kompsat-2 satellite image was used for forest canopy gaps delineation (a). The classification was based on an artificial neural network and allowed the identification of 297 canopy gaps (50-1776 m(2)). The size and spatial pattern of the gaps were found to be different between core area and buffer zone. Forest structure (b) in the core area was analysed applying a 120 m grid sampling design (40 plots). From this preliminary analysis we identified a stand with typical old-growth characteristics in the central part of the core area and a 1 ha intensive sampling plot was established (c). Inside the plot all trees were mapped, measured, and a core extracted to assess tree age. Spatial structure was investigated by means of Point Pattern Analysis and LISA (Local Indices of Spatial Autocorrelation). The three scales and approaches produced consistent results. The landscape approach confirmed the hypothesis that small-scale processes predominate at Lom especially within the core area of the reserve. At the forest stand level the presence of small scale disturbance/mortality processes was confirmed. (C) 2010 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of spatial Statistics 2011