Directionality works but does not scale, and inherited adaptation does not appear, in a resource-limited pond world

A pond world with a resource, death and reproduction is used to ask two questions. First, does a local behavioural rule help after the environment changes? Local resource-greedy movement does: it holds 0.475 of the favourable half against 0.357 for the control on a patchy environment. The advantage decays as the rich zone fills (correlation -0.57 with rich-zone occupancy over 18 points), because cells compete for space rather than for food: one organism per grid cell, and the patch's inflow exceeds one cell's consumption, so the resource barely falls while occupancy rises. Second, is such a skill inherited? Four tests -- a double flip back to the native environment, a flip to a new environment, a heritable mobility trait and a heritable directionality trait, each with a transplant control -- give a consistent null, and the null survives the removal of the lineage-monopoly confound. Three experiments are recorded as invalid by construction rather than as negative results. Every printed number is re-derived from the deposited artifacts by a machine check (49 checks, 0 failures).