Off the rocky coast of Jervis Bay, Australia, an octopus scoops a fistful of silt into its arms, angles its siphon, and blasts the debris straight at a neighbor sitting a few body lengths away. The neighbor flinches, raising its arms as the cloud of sediment arrives. Seconds later, the thrower reloads and fires again.
That scene, captured on underwater cameras anchored to the seafloor inside Booderee National Park, is one of dozens of targeted throws documented in a peer-reviewed study published in PLOS ONE in November 2022. Led by philosopher of science and marine biologist Peter Godfrey-Smith of the University of Sydney, the research team recorded wild gloomy octopuses (Octopus tetricus) collecting shells, silt, and algae, then launching the material at other octopuses using their siphon, the muscular tube these animals normally use for jet propulsion. As of June 2026, the paper remains the first quantified, peer-reviewed evidence that any octopus species throws objects during social encounters and sometimes connects with its target.
What the cameras recorded
The team deployed stationary cameras at a site where an unusually dense cluster of gloomy octopuses had built dens on the same small patch of temperate reef. Across sampling periods in 2015 and 2016, the footage captured 102 throws that met the researchers’ strict operational definition: an octopus had to gather material, hold it in its arms, and propel it outward through a visible siphon jet. Incidental sediment displacement, the kind that happens when an octopus simply moves across the bottom, did not count.
Not all throws looked the same. Many occurred when an octopus was housekeeping, clearing shells or food scraps from its den. These tended to be lower in force and less directional. But throws that happened during direct social encounters, when a neighbor approached, displayed, or attempted to mate, told a different story. They were more forceful on average, more likely to use silt as ammunition rather than shells, and more often sent debris along a path that passed close to or struck another individual. The throwers also changed color during these social throws, shifting to a uniform dark pattern that the researchers interpreted as a possible signal of heightened arousal or aggression.
Five detailed video clips published in the study’s supplementary materials show specific throws frame by frame, documenting the type of debris used, the vigor of each launch, the thrower’s body pattern, and whether the projectile hit its mark. In several sequences, a single octopus threw multiple times in quick succession, adjusting its body orientation between volleys in a way that appeared to track a neighbor’s position.
The cameras also caught the reactions of octopuses on the receiving end. Some raised their arms defensively. Others ducked or shifted position the moment debris arrived. In at least one case, a target octopus altered its behavior after being hit, moving away from the thrower. Those responses do not prove the thrower intended to provoke them, but they show the projectiles had real, observable effects on nearby animals.
Why intent is the hardest question
Projectile use among nonhuman animals is rare. Chimpanzees throw stones. Archerfish spit jets of water to knock insects off branches. But confirmed cases of one animal directing a projectile at another member of its own species are rarer still, and proving that the thrower meant to hit its target is a problem that haunts every study in this space.
As a Nature news report on the findings noted, researchers cannot ask an octopus what it intended. Godfrey-Smith and his co-authors addressed the gap with statistics: throws during social interactions differed from non-social throws in force, ammunition choice, body color, and trajectory in ways that were consistent with deliberate targeting. That is strong circumstantial evidence. It is not, however, a direct window into octopus cognition, and the authors were careful to frame it that way.
Whether the throwing represents territorial defense, a form of social signaling, simple irritation at a too-close neighbor, or some mix of all three is not yet resolved. The study gives future researchers a scoring framework and a set of testable predictions, but it does not claim to have cracked the question of octopus intent.
A population quirk or a species-wide behavior?
The Booderee site is unusual. Gloomy octopuses are generally solitary, but at this location, a concentration of suitable den sites on a small reef patch forces many individuals into close quarters. That density creates frequent social encounters of the kind that rarely occur in more typical, spread-out habitats.
It remains an open question whether octopuses living at lower densities, or in different environments, would throw debris as often or in the same social contexts. No comparable video dataset from another location has been published. Until follow-up fieldwork is done elsewhere, it is possible that the throwing behavior documented at Booderee is partly a product of local crowding rather than a universal trait of the species.
That uncertainty does not diminish what the study established. Wild Octopus tetricus physically propel collected debris through siphon jets, these events occur during social encounters, and the debris sometimes hits other octopuses. Those three facts are new to the scientific record and have already shifted how biologists think about cephalopod social lives.
What this means for octopus science
For decades, the default view of octopuses cast them as brilliant but antisocial: sophisticated problem-solvers that interact with objects mainly to build shelters or crack open prey. The Booderee footage complicates that picture. If octopuses can gather ammunition, aim it at a specific individual, and adjust their throws in real time, their social toolkit is more elaborate than their loner reputation suggests.
The study also hands other research teams a ready-made template. Its operational definitions for “throw” and “hit,” its frame-by-frame scoring method, and its statistical approach to separating social from non-social throws can all be applied at new sites and with other cephalopod species. The most valuable next step, as Godfrey-Smith and colleagues acknowledged, is replication: more cameras, more locations, more species. Only then will scientists know whether the shell-hurling octopuses of Jervis Bay are outliers or the first confirmed example of a behavior hiding in plain sight across the world’s oceans.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
