Patch size and distance: modelling habitat structure from the perspective of clonal growth

This study considers the spatial structure of patchy habitats from the perspective of plants that forage for resources by clonal growth. Modelling is used in order to compare two basic strategies, which differ in the response of the plant to a patch boundary. The 'avoiding plant' (A) never grows out of a good (resource-rich) patch into a bad (resource-poor) region, because the parent ramet withdraws its subsidy from the offspring. The 'entering plant' (E) always crosses the boundary, as the offspring is subsidized at the expense of the parent. In addition to these two extreme scenarios, an intermediate mixed strategy (M) will also be tested. The model is used to compare the efficiency of foraging in various habitats in which the proportion of resource-rich areas (p) is varied.

A stochastic cellular automata (CA) model is developed in which habitat space is represented by a honeycomb lattice. Each cell within the lattice can accommodate a single ramet, and colonization can occur from a parent ramet's cell into six neighbouring cells. The CA consists of two layers: the population layer and the habitat. In the population layer, a cell can be empty or occupied by a ramet; in the habitat layer, a cell can be good (resource-rich) or bad (resource-poor). The habitat layer is constant; the population layer changes over time, according to the birth and death of ramets.

Strategies M and E are primarily limited by patch distance, whereas A is more sensitive to patch size. At a critical threshold of the proportion of resource-rich areas, p = 0·5, the mean patch size increases abruptly. Below the threshold, E is more efficient than A, whilst above the threshold the opposite is true. The mixed strategy (M) is more efficient than either of the pure strategies across a broad range of p values.

The model predicts more species/genotypes with the 'entering' strategy, E, in habitats where resource-rich patches are scattered, and more plants with the 'avoiding' strategy, A, in habitats where the connectivity of resource-rich patches is high. The results suggest that the degree of physiological integration between a parent and an offspring ramet is important even across a very short distance because it can strongly influence the efficiency of foraging.