Some clonal plant species decrease rhizome or stolen internode lengths
and/or increase the frequency of branching when they grow in favorabl
e environments. This ''foraging response'' is thought to be beneficial
since it should allow ramets to concentrate in areas of favorable hab
itat. However, there have been few critical tests of the effectiveness
with which ramets are placed in favorable habitat as a result of the
foraging response. In this paper, I use empirically calibrated stochas
tic simulation and diffusion models to compare the growth of clones in
favorable and unfavorable habitat. I ask whether observed changes in
rhizome lengths and clonal branching patterns are likely to decrease s
ignificantly the distance clones move, and thus, to enable ramets to r
emain for longer periods of time in favorable habitat. For the empiric
al data used in this study, results from the models indicate that the
effectiveness of the foraging response is likely to be variable. In so
me cases, such as results from models based on the response of Glechom
a hederacea to nutrients, there is no significant difference in the di
stance clones move in favorable and unfavorable habitat. Thus, even th
ough rhizome lengths may be significantly shorter in favorable patches
, this does not guarantee that clones disperse significantly less far
and thereby remain longer in favorable habitat. The effectiveness of t
he foraging response depends strongly on the distribution of clonal gr
owth angles, the pattern of clonal branching, and the variance in rhiz
ome or stolen internode length. These results, particularly the import
ance of growth angles and the relatively limited effectiveness of the
foraging response, differ from those in previously published models of
foraging in clonal plant species. I conclude with a discussion of rea
sons that may underlie these differences.