Bottlenose dolphins in Shark Bay, Australia, break off a cone-shaped marine sponge and wear it over the rostrum ("beak") as a glove while probing rocky, rubble-strewn channel floor for fish — a genuine, matrilineally-transmitted example of tool use in a wild mammal. The behavior is not evenly spread: it is passed almost exclusively from mother to (mostly female) calf through years of close observation.
This model runs two things at once. A fast background Monte-Carlo population (thousands of probes) gives statistically stable success-rate and energy-per-time numbers for "sponge-equipped" vs. "bare-rostrum" foraging strategies. Two visible agents — a sponge-equipped mother and a naive juvenile — forage on the grid in real time; every time the mother successfully extracts a fish with her sponge while the juvenile is nearby, there is a chance (the learning-probability slider) the juvenile adopts sponging for the rest of its life.
- Smooth sand — easy to probe, moderate fish density, available to any dolphin regardless of tool use.
- Rocky rubble — hides more, often larger fish (e.g. barred sandperch), but a bare rostrum probing it has both a low success chance and a real injury risk (sharp coral, urchins, stonefish spines). A sponge blunts that risk and lets the animal probe as effectively as on sand.
- Bare-rostrum dolphins mostly avoid rubble rather than risk it every time — exactly as observed in Shark Bay, where non-sponging dolphins forage predominantly over sand/seagrass and rarely enter the deep rocky channels spongers specialize in.
- Net energy/time is (energy gained from successful catches − energy spent probing and carrying) divided by elapsed foraging time — the standard optimal-foraging currency for comparing strategies.