ABSTRACT: A few studies have shown that sponges are capable of locomotion on the substrata. However, the ecological significance of this ability remains virtually unexplored to date. Here, we conduct a series of experiments documenting the movementpatterns of the encrusting sponge Scopalina lophyropoda as a function of food availability, density of neighbors, water flow, burial by silt, and body size. In all experiments, we used small sponges that were developed from explants afterattachment to acetate sheets that served as substrata. The first experiment revealed that starvation is a powerful stimulus to locomotion, since starved sponges displaced over significantly greater distances than did fed sponges, irrespective of thedensity of neighbors. It was also found that paths were significantly more convoluted when sponges moved through high-density groups, irrespective of food availability. In a second experiment, we found that movement of an isolated sponge in the vicinityof barriers made of either histocompatible (isogeneic) or histoincompatible (allogeneic) conspecifics was not random, but was oriented significantly towards the neighboring barrier. Sponges were unable to determine neighbor compatibility before physicalcontact was made. In all isogeneic encounters, the outcome was tissue fusion. In contrast, all allogeneic encounters were characterized by an initial histological rejection and a subsequent change in path direction to avoid a situation of prolongedcontact between sponges. In a third experiment, we found that sponges under conditions of dispersed flow with silt deposition traveled significantly further than sponges under directed flow without silt deposition. Direction of movement was random underconditions of dispersed flow, but sponges under conditions of directed flow moved perpendicularly to the flow direction, avoiding contact with the high-speed water jets. In a fourth field experiment, we assessed differences in mobility between relativelylarge and small individuals (i.e. in a size range equivalent to that of 1 yr old and 1 wk old postlarvae, respectively). Mobility was similar in both size ranges. The results of the experiments strongly suggest that sponges in the field may use thecapability of locomotion to prevent prolonged physical contact with competitors and to move away from sites with limited access to food, or excessive exposure to silt and abrasion by water flows. The histological study revealed that moving sponges mayincur certain costs, since both the canal system for water pumping and the skeleton responsible for the attachment of the sponge to the substratum have to be continuously reorganized as sponges move. Nevertheless, the selective (ecological) benefitsderived from this ability must offset, or at least be in step with, histological and energetic costs. Were this not the case, sponges would not have evolved their capability of locomotion.