Spatial organisation of a bimodal forest stand

Many populations have a bimodal size distribution, even when composed of a single cohort. In developing forest stands, this pattern is usually attributed to asymmetric competition at canopy closure among trees which have access to the upper canopy and those which have failed to reach it. Nevertheless, alternative explanations for bimodality exist, and in sessile organisms spatial pattern analysis can be used to compare their predictions. A 0. 25-ha plot was created in a maturing stand of Asiatic white birch (Betula platyphylla Sukacz.) in Central Kamchatka. All stems >1 cm diameter at breast height (DBH) were fully mapped. Mark correlation analysis revealed size compensation among stems up to 3.5 m apart, providing evidence that competition affected the distribution of stem sizes. The spatial pattern of trees was analysed using the pair correlation function g(r). Large trees (>20 cm DBH) had a dispersed distribution to which a Strauss soft-core Gibbs process model was fitted. This suggested that large trees interacted at scales up to 4.16 m. Small trees (1-20 cm DBH) were distributed randomly, but a cross-pair correlation analysis revealed a greater likelihood of occurrence beginning at 4.3 m from large trees, closely matching the modelled interaction distance. These results are consistent with the hypothesis that asymmetric competition is structuring this bimodal cohort of trees: large stems tend towards a dispersed pattern, exerting competitive effects at scales up to approximately 4 m, whereas smaller stems are more commonly found in the interstices within the pattern of large trees.