Abstract
Aims This paper reviews the way optimization theory has been used in canopy models to analyse the adaptive significance of photosynthesis-related plant characteristics and their consequences for the structure and species composition of vegetation stands. • Scope In most studies simple optimization has been used with trait values optimal when they lead to maximum whole-stand photosynthesis. This approach is subject to the condition that the optimum for one individual is independent of the characteristics of its neighbours. This seems unlikely in vegetation stands where neighbour plants strongly influence each other's light climate. Not surprisingly, there are consistent deviations between predicted plant traits and real values: plants tend to be taller, distribute nitrogen more evenly among their leaves and produce more leaf area which is projected more horizontally than predicted by models. • Conclusions By applying game theory to individual plant-based canopy models, other studies have shown that optimal vegetation stands with maximum whole-stand photosynthesis are not evolutionarily stable. They can be successfully invaded by mutants that are taller, project their leaves more horizontally or that produce greater than optimal leaf areas. While these individual-based models can successfully predict the canopy structure of vegetation stands, they are invariably determined at unique optimal trait values. They do not allow for the co-existence of more than one species with different characteristics. Canopy models can contribute to our understanding of species coexistence through (a) simultaneous analysis of the various traits that determine light capture and photosynthesis and the trade-offs between them, and (b) consideration of trade-offs associated with specialization to different positions in the niche space defined by temporal and spatial heterogeneity of resources.