An Architectural Approach to the Comparative Ecology of Plant Root Systems

S ummary Although plants devote a large proportion of their resources to roots, we have a poor understanding of the constraints under which root systems function. Roots are much less variable morphologically than leaves and it is likely that root systems rather than individual roots are the focus of natural selection. In other words, architecture is more important than morphology. Existing classifications of root systems, based on the developmental model, have failed to provide much insight into their functioning and an alternative, topological model is outlined, in which the link is the basic unit of classification. Other components of the architecture of root systems, including link lengths, branching angles and diameters, are considered and the ecological implications of variation in each is discussed. Simulation models of transport and space exploration are discussed and it is shown that resource cost, transport efficiency and exploration efficiency cannot be simultaneously minimized and that optimum form may vary with the mobility of the resource. In general, a ‘herringbone’ structure seems to be the most efficient at exploration of space but the least transport‐efficient and the most expensive.