Morning Overview

Octopuses and fish were just filmed hunting as a team on the reef — and the octopus punches any partner that slacks off, keeping the whole group in line

Off the coast of a coral reef in the Red Sea, an octopus glides across a patch of rubble, eight arms probing every crack. A small entourage of reef fish follows close behind: goatfish hovering near the substrate, a blacktip grouper hanging just above, a tailspot wrasse darting at the edges. The group moves like a single organism with no obvious leader. Then one of the trailing fish drifts too far from the action, apparently freeloading. The octopus whips an arm sideways and smacks the fish so hard it tumbles backward through the water column. Seconds later, the group re-forms and keeps hunting.

That scene, captured on underwater video and published in a 2024 study in Nature Ecology & Evolution, is part of a growing body of evidence that octopus-fish hunting partnerships are far more organized, and far more ruthlessly enforced, than scientists previously understood. The research, led by marine biologist Eduardo Sampaio of the Max Planck Institute of Animal Behavior and the University of Lisbon, used three-dimensional tracking and behavioral coding to reveal that these mixed-species groups share leadership, adjust their composition on the fly, and rely on physical punishment to keep partners in line.

A hunting alliance with teeth

The study documented groups of day octopus (Octopus cyanea) hunting alongside several reef fish species on coral reefs. Using underwater video paired with 3D spatial reconstruction, Sampaio and colleagues tracked how influence shifted within each group. No single species permanently controlled direction or pace. In some sequences, the octopus initiated a foraging run and fish fell in behind. In others, fish surged ahead to investigate crevices or chase fleeing prey while the octopus lagged, only to retake the lead moments later.

The logic of the partnership is straightforward. Octopuses are superb at extracting prey from tight spaces, reaching into holes and crevices that fish cannot access. Fish, in turn, are faster in open water and can flush prey from exposed positions back toward the octopus. Together, the group covers more ground and more habitat types than either party could alone.

But the most striking finding was the punching. The researchers defined it precisely: an explosive, directed arm motion that physically displaces a fish from its position within the group. Supplementary video published alongside the paper shows clear examples of octopuses striking fish and sending them tumbling. The underlying tracking data, positional coordinates, time stamps, and behavioral annotations are publicly available through a figshare repository, allowing independent verification.

Sampaio has described the punching in interviews as potentially serving multiple functions: punishing freeloaders who take prey without contributing, repositioning partners for tactical advantage, or simply asserting dominance over a contested catch. The study itself does not claim to resolve which motivation dominates in every case. In some clips, the octopus appears to strike a fish that just intercepted prey. In others, the target seems to be a bystander that has not obviously broken any rules.

Decades of evidence across multiple oceans

These partnerships are not a new discovery. Field researchers have documented octopus-fish hunting alliances across multiple ocean basins for decades, though the sophistication of the coordination was not fully appreciated until recently.

In the Red Sea’s Gulf of Eilat, a 1985 study observed groupers associating with what researchers called “nuclear” predators, either an octopus or a moray eel, to exploit prey flushed by the lead hunter. Fish that joined these coalitions gained access to prey they could not easily dislodge on their own. On a remote Indian Ocean reef, researchers recorded multispecies hunting associations involving day octopus and several reef fish species acting in concert, reinforcing the idea that such mixed groups are a recurring feature of tropical reef ecosystems rather than a local oddity.

In French Polynesia, field studies in the 1990s documented Octopus cyanea foraging on coral atolls, providing baseline behavioral data that later work built upon. Those observations showed that O. cyanea regularly patrols complex reef structures, probing holes and overhangs where fish and crustaceans hide. When fish accompany these patrols, they capitalize on prey fleeing the octopus, while the octopus may benefit from fish flushing targets from open water into crevices it can reach.

Perhaps the most remarkable earlier finding involves cross-species communication. A 2013 study published in Nature Communications showed that coral trout (a type of grouper) use a referential pointing signal to indicate hidden prey to cooperative hunting partners, including Octopus cyanea. The grouper positions its body vertically and directs its head toward concealed prey, sometimes combining the posture with rapid head shakes. This effectively guides the octopus to a target it cannot see on its own. At the time, referential gesturing between species was thought to be limited to primates and corvids, so finding it on a coral reef was a genuine surprise.

What scientists still do not know

The new study establishes that punching happens and that it displaces fish, but several important questions remain open as of June 2026.

Long-term tracking of individual fish that have been punched is absent from both the new paper and the historical studies. Whether a punched fish leaves the group permanently, returns later, or simply repositions and continues hunting has not been documented over extended time periods. Without that data, it is difficult to say whether punching functions as a one-time correction or a sustained enforcement mechanism that shapes group membership over days or weeks. The same individual fish may be hit repeatedly, or octopuses may preferentially target newcomers; for now, those possibilities remain untested.

The quantitative success rates by group composition reported in the study rest entirely on the original authors’ statistical framework. Peer review provides a layer of confidence, but no independent team has yet reanalyzed the raw data or attempted replication. Alternative analytical choices, such as different thresholds for defining leadership or different ways of scoring partial prey captures, could shift the strength of the reported effects.

Environmental context matters too. The exact reef sites for the new video observations are tied to supplementary materials rather than publicized with precise geographic coordinates, which limits the ability of other teams to revisit the same locations. Local predator density, human disturbance, water temperature, and reef complexity could all influence how often punching occurs or which species participate. Until comparable datasets are collected from other reefs using similar 3D tracking methods, it will be hard to know how representative the documented behavior truly is.

Why a punch on a reef changes how we think about cooperation

What makes this research significant is not just the spectacle of an octopus slugging a fish. It is what the behavior implies about the architecture of cooperation between species that share no common ancestry, no kinship, and no language.

Most examples of complex cooperation in the animal kingdom involve members of the same species: wolves in a pack, dolphins in a pod, chimpanzees on a border patrol. When cooperation crosses species lines, it tends to be simple and passive, like the oxpecker that picks ticks off a buffalo. The octopus-fish hunting groups documented by Sampaio and colleagues are something different. Leadership is shared and fluid. Group composition changes. And when a partner fails to contribute, there is a physical consequence.

If fish can direct octopuses toward hidden prey using referential gestures, and octopuses can enforce participation with targeted strikes, then the relationship between these species is far more negotiated than simple opportunistic following. Both parties appear to exchange information and adjust their behavior based on what the other does. Punching fits into this picture as one more tool in a surprisingly complex toolkit for managing cooperation without shared biology or stable social hierarchies. On a reef where every missed meal matters, a well-placed arm strike may be the bluntest, and most effective, way to keep the team together.

More from Morning Overview

*This article was researched with the help of AI, with human editors creating the final content.