Fish can use tools. Not just use them in a reflexive, scripted way, but select them, transport them, test them for fitness, and then deploy them with precision. This shouldn't be possible. Until 2009, it wasn't supposed to happen at all.
Most people still think tool use is a primate thing, maybe a crow thing if they've seen a nature documentary. The assumption runs deep: you need hands, or at least sophisticated manipulative digits. You need a big brain housed in a skull. You definitely don't need gills. Yet here we are, watching fish that weigh less than a pound execute multi-step plans that would make an engineer nod approvingly. A tuskfish, according to research documented since the early 2000s, will dig a clam from the sand, carry it in its mouth to a nearby rock, position itself carefully, and then repeatedly throw the clam against the rock until the shell cracks. Then it eats the meat inside. The process requires forward planning, transport logistics, and what researchers call "testing behavior"—the fish literally inspects the clam and the rock before committing to the work.
The initial discovery in 2009 shocked the scientific community enough that it prompted serious study. What emerged over the following years was something more unsettling: tuskfish weren't anomalies. They were just the first fish we'd caught doing this. According to recent 2025 research, at least three wrasse species have now been documented using rocks as anvils in similar ways, selecting specific stones and transporting them to feeding sites. The behavior is consistent. It's learned. And it requires the kind of cognitive scaffolding—do this, then that, then verify the result—that we usually reserve for animals we consider genuinely intelligent.
The mechanism here is still being unpacked, but the leading theory centers on what researchers call "ecological opportunity." Fish with access to hard-shelled prey but without the dental apparatus to crack them open face a genuine problem: how do you access a high-calorie food source when your mouth can't do the job? Tool use solves that problem, but only if the fish can imagine the solution first. Only if it can remember where rocks are, retrieve them, and understand that smashing creates results. This isn't instinct in the traditional sense. Baby tuskfish don't arrive knowing how to do this. They learn it, sometimes by watching other fish, sometimes through trial and error. The behavior itself becomes culturally embedded in populations—some groups of fish use tools, others in the same species don't.
What's genuinely strange is what this implies about fish cognition more broadly. If a creature with a brain the size of a pea can plan, test, and execute a tool-use strategy, our entire framework for measuring animal intelligence starts to wobble. We've spent decades building hierarchies of cognition with fish somewhere near the bottom, treated as reactive creatures with three-second memories. The evidence suggests we've been wrong—not just slightly wrong, but structurally wrong about how to think about minds in the animal kingdom. The tuskfish isn't exceptional because it's smarter than a fish. It's exceptional because it's a fish at all.