Forest structure in space: a case study of an old growth spruce-®r forest in Changbaishan Natural Reserve, PR China

Ecologists and foresters are becoming increasingly aware of the importance of spatial information in ecosystem analysis and resource management. An across-scale analysis of forest structure was conducted to characterize the spatial characteristics of a 2 ha spruce-®r forest located inside Changbaishan Natural Reserve (CNR), PR China. The study was designed to develop an approach for assessment of within-stand heterogeneity to increase understanding of the effects of heterogeneity on pattern± process relationships in forests. Univariate and bivariate Ripley's K functions were employed to capture stand heterogeneity in terms of intraand inter-speci®c point patterns of tree distributions. Stem-mapped crowns were generated and analyzed as canopy patches within a geographic information system (GIS) to quantify patterns within the strata of the forest canopy. The dominant spruces were randomly distributed within each height class but aggregated when all height classes were analyzed. In addition, spruces had repulsive patterns to other species which exhibited clear aggregation. Canopy heterogeneity, both in terms of composition and spatial patterning, were complex ± as expressed within canopy layers and among individual forest species. This across-scale complexity supports a multivariate spatial and across-scale approach to characterizing forest structure, and argues for linked pattern-process experiments. We suggest that joint ®eld and simulation studies be conducted which relate changes in forest stand dynamics to changes in stand heterogeneity. These linked studies are needed to provide a measure of ecological signi®cance relative to statistical signi®cance of patterns. # 1999 Elsevier Science B.V. All rights