Tree spacing and area of competitive influence do not scale with tree size in an African rain forest.

Questions: Is the area of influence of individual trees determined by tree size? Does competition, inferred from spatial pattern between neighbouring trees, affect adult tree spacing patterns in an tropical forest? At what size-class or stage is competition between neighbours most likely to affect adult tree spacing patterns? Location: Kibale National Park, western Uganda. Methods: Relationships between focal tree size and nearest neighbour distance, size, density, and species in a 4-ha permanent plot, using point pattern analyses. Results: We found non-random patterns of distribution of nearest tree neighbours (stems > 10 cm DBH). Independent of identity, tree density was highest and neighbours were regularly spaced within 3-5m of an individual. Tree densities were lower and relatively constant at distances >5m and neighbours were typically randomly spaced. In general, conspecific patterns conformed to the latter trends. Thus, individual area of influence was small (within a radius of 3-5 m). Rarer species were more clumped than common species. Weak competitive thinning occurred among more densely packed small trees (<20 cm DBH), and rapidly disappeared with increasing tree size and distance from an individual. The clumping and density of individuals was not significantly affected by tree size. Conclusions: Negative effects of competition among trees are weak, occur within the crown radius of most individuals, and are independent of adult tree size and identity. The density of neighbouring trees (aggregation) did not decline with increasing focal tree size at either the conspecific or the community level and tree diameter (tree size) was not a good estimator of the implied competitive influence of a tree. Mechanisms operating at the recruitment stage may be important determinants of adult tree community diversity and spacing patterns. [ABSTRACT FROM AUTHOR]