A self‐assembling model of resource dynamics and plant growth incorporating plant functional types

Summary We explored the ecological properties of a self‐assembling, cellular automaton model by simulating the behaviour of distinct plant functional types. Types differed according to three fundamental traits: size of plant modules; longevity of modules in the absence of resources; and propensity to flowering. When expressed at three levels apiece, the three traits were able to approximate the essential biology of seven important types within a well known system of plant functional types (the C‐S‐R system, based on disturbance and resource availability). The seven types, when grown in real experiments with virtual communities, reproduced classical community‐level behaviour to the extent of demonstrating smooth, competitive replacements along gradients of resource or disturbance, or both. In complex communities simulated over a wide range of environmental conditions, the seven types produced a clearly ‘humpbacked’ curve linking plant diversity to community productivity. The mixtures of species resulting from these experiments together represented all parts of C‐S‐R space. These outcomes emerged solely from the ‘bottom‐up’ processes inherent in self‐assembling cellular automata. No community‐level specifications were inserted into the model, yet the community‐level consequences agreed with both the underlying predictions of C‐S‐R theory and field observations. This work confirms that the dynamics of herbaceous plant communities can be emergent from the resource‐handling properties of the organs of the constituent species, and that a simple rule base can, to a first approximation, be sufficient to reproduce such dynamics.