Mechanisms of homing in the fiddler crab Uca rapax. 2. Information sources and frame of reference for a path integration system.

Fiddler crabs Uca rapax are central-place foragers, making feeding excursions of up to several meters from their burrows. This study investigates the sources of directional and distance information used by these crabs when returning to their burrows. We tested the spatial frame of reference (egocentric or exocentric), and the source of spatial information (idiothetic or allothetic) used during homing. We also tested which components of their locomotion they integrated (only voluntary, or voluntary plus reflexive). Fiddler crabs in their natural mudflat habitat were passively rotated during normal foraging behavior using experimenter-controlled disks, before they returned home. Crabs resisted passive rotations on the disk by counter-rotating when the disk turned, which was a compensatory response to unintended movement. Crabs were usually situated eccentrically on the disk, and therefore were also subjected to a translation when the disk rotated. No crab actively compensated for this translation. Crabs that fully compensated for disk rotation made no directional homing error. Crabs that did not fully compensate homed in a direction that reflected their new body orientation. In other words, if we succeeded in reorienting a crab (i.e. it undercompensated for disk rotation), its homing error was equal to the angle by which it had been reoriented, regardless of the magnitude of the optomotor compensation. Computer-modelled crabs, each equipped with a path integrator utilizing different combinations of external (allothetic) and path-related (idiothetic) input, traversed the digitized paths of the real crabs. The home vector computed by the model crab was then compared to the homing direction observed in the real crab. The model home vector that most closely matched that of the real crab was taken to comprise the path integration mechanism employed by fiddler crabs. The model that best matched the real crab gained direction and distance idiothetically (from internal sources such as proprioceptors), and integrated only voluntary locomotory information. Crabs were also made to run home across a patch of wet acetate, on which they slipped and were thus forced to take more steps on the homeward path than theoretically required by the home vector. Crabs whose running velocity across the patch was unusually low also stopped short of their burrow before finding it. Crabs whose running velocity was not impeded by the patch did not stop short, but ran straight to the burrow entrance, as did control crabs that ran home with no slippery patch. We interpret this to mean that the velocity of some crabs was impeded because of slipping, and these therefore stopped short of their burrow after having run out their homing vector. This is positive evidence in support of the hypothesis that path integration is mediated either by leg proprioceptors or by efferent commands, but our data do not allow us to distinguish between these two possibilities.