COMPETITIVE ASYMMETRY REDUCES SPATIAL EFFECTS ON SIZE-STRUCTURE DYNAMICS IN PLANT-POPULATIONS

The growth of each individual in plant populations was simulated by a spatial competition model for five density levels and four different spatial distribution patterns of individuals, varying from highly clumped to regular. The simulation results were analysed using the diffusion model for evaluating the effects of density and distribution pattern on the size-structure dynamics in relation to the degree of competitive asymmetry. At low densities, changes in statistics of plant weight over time such as mean, coefficient of variation, skewness, and Box-Cox-transformed kurtosis differed greatly among spatial patterns, irrespective of the degree of competitive asymmetry. In completely symmetric competition, the spatial effect on size-structure dynamics remained relatively large irrespective of densities, although mean plant weight became similar among the spatial patterns with increasing density. However, the spatial effect diminished with increased density in strongly asymmetric competition, when similar size distributions were realized irrespective of the spatial patterns. Therefore, it was concluded that: (1) irrespective of the degree of competitive asymmetry, spatial pattern is important for size-structure dynamics at low densities; (2) spatial pattern is nearly immaterial under strongly asymmetric competition at high densities; and (3) under crowded conditions, neighbourhood effects are much more apparent at the population level in less asymmetric competition. These processes and outcomes are linked to the forms of the functions of mean growth rate of individuals [G(t,x) function] and variance in growth rate [D(t,x) function]. These functions are variable depending on the spatial pattern under symmetric competition, but are rather stable under strongly asymmetric competition at high densities irrespective of the spatial patterns. Therefore, size structure under strongly asymmetric competition can be regarded as a stable system, whereas that under symmetric competition is regarded as a variable system in relation to the spatial pattern and process. From this, it was inferred that: (1) the goodness-of-fit of spatial competition models for crowded plant populations is higher in less asymmetric competition; and (2) higher species diversity in plant communities is associated with the lower degree of competitive asymmetry.