Simple mechanisms of size distribution dynamics in crowded and uncrowded virtual monocultures

Size structure dynamics in even-aged monocultures sometimes exhibits fairly complex patterns. Likewise, models of competition among individuals are also able to produce various changes in size distributions. This raises the question about the reliability of inferences from patterns (size distributions) to processes occurring in plant monocultures. We review some possible mechanisms of size distribution dynamics and their pure, isolated effects identified in simple models of growth and competition without introducing mortality into the models. Variability in growth patterns may produce marked nonmonotonic changes in size distributions. Individual differences in growth parameters imply differential responses of individuals to competition, without introducing mortality into the models. Positive correlations between these parameters result in trade-offs between fast growth in uncrowded conditions and relative insensitivity to crowding, i.e. the ability to grow under crowded conditions. Under some circumstances, these trade-offs may cause symmetric competition to promote an increase in variation and mass distribution skewness with increasing density and reduce mass order reversals during growth—both effects usually attributed to asymmetric competition. In turn, under asymmetric competition variation and skewness increase with density up to some point only and then decrease. The local nature of competition brings about a variety of distribution changes depending on population density. These spatial effects are sensitive to other factors in the case of symmetric competition, but to a lesser extent in the case of asymmetric competition. Competitive asymmetry reduces easily detectable spatial effects. The magnitude of this reduction depends on the amount of initial mass variation. Based on the relative complexity of these results, we conclude that reliable inferences from size structures are not possible. We also point to some consequences of model behavior complexity for the analysis of simulation models. Presenting only selected patterns generated by a model may be as misleading as trying to infer process from a simple pattern.