Orcas off the west coast of North America are grooming each other with kelp, in a rare sighting of marine mammals manufacturing and using tools.

For several years, scientists have been keenly observing 80 endangered killer whales in the segment of the Pacific Ocean between British Columbia and Washington state. To get a bird鈥檚 eye view of the whales鈥� lives, the researchers also tracked them with drones.

While poring over footage from summer 2024, they noticed that the orcas were manoeuvring strands of kelp in odd ways. It was 鈥渞eally weird鈥�, says at the in Washington state, 鈥渂ut the whales, they do weird things all the time鈥�.

In the footage, the orcas can be spotted breaking off kelp stalks near where they meet the rock bed by grabbing them with their teeth and jerking their heads back and forth. The short, snapped-off segments were roughly equivalent in length to that of the whale鈥檚 beak-like face. Over and over again, the orcas appear to consistently target just that specific segment of the algae, not other random parts of kelp.

After breaking off a strand, a whale would then sandwich the kelp between their head and the bodies of other whales in the pod, rubbing and rolling it onto each other鈥檚 sides. They take turns cleaning each other with the kelp, sometimes grooming each other for up to 12 minutes.

鈥淲hat鈥檚 cool is that they don鈥檛 have any kind of hand-like appendages, and so they鈥檙e doing all of this using very deliberate movements of their body,鈥� says John.

Free newsletter

Sign up to The Earth Edition

Sign up to newsletter

Orcas are known to rub themselves against kelp on their own, known as 鈥渒elping鈥�. This could be a social variation of that behaviour. 鈥淲e know that the social bonds in this population are super, super strong, and we know that contact is one way that they reinforce those bonds,鈥� says John.

The behaviour was present across all ages and sexes, though the data suggests the whales that were most closely related and those closer in age were more likely to 鈥渒elp鈥� together. Crucially, this may be a form of whale hygiene, says John, as the team found that orcas are more likely to scrub each other with kelp if they are shedding their skin.

It might still be too early to confirm whether this skincare has health benefits, says at Griffith University, Australia. He would like to see the researchers cross-reference the orcas鈥� skin bacteria with the properties of the kelp to see if they match.

鈥淚t totally makes sense to me that they are seeking out anything that the ocean could offer to help them with potential reduction in skin infections,鈥� says Meynecke. He suspects this is a widespread behaviour among other orca populations and whale species.

Budgerigars are some of the chattiest birds, and that is reflected in their brains. Budgie brains contain a map of vocal sounds, which is similar to that found in the human brain and has not been seen in any other bird.

鈥淲e found that there was a representation of vocal sounds in a part of the brain that is analogous to a key speech region in the human brain,鈥� says at the聽.

Budgerigars (Melopsittacus undulatus), also known as parakeets, are small parrots native to Australia. They are spectacular vocal learners, able to mimic a variety of sounds including human speech. A budgie called Puck had a vocabulary of about 1728 words, . 鈥淭he ability to vocally imitate is something that is extremely rare in the animal kingdom,鈥� says Long.

With also at NYU鈥檚 medical school, Long used silicon probes to record the electrical activity in budgies鈥� brains. They focused on part of the forebrain, the central nucleus of the anterior arcopallium, which was known to be involved in the motor control of vocalisations. As the budgies made calls, Long and Yang tracked how their electrical activity changed.

鈥淥ur study was the first to measure the activity in the parrot brain during vocalisation,鈥� says Long.

The pair found neurons in the central nucleus of the anterior arcopallium that were active only when the budgies made specific sounds. 鈥淭here are cells that are active for consonants,鈥� says Long. Others do vowels, with some active for high-pitched sounds and others for low-pitched.

Free newsletter

Sign up to The Earth Edition

Sign up to newsletter

Long compares this brain structure to a keyboard. 鈥淚t has this kind of set of keys, or in this case, set of brain cells, that can represent each one of these vocal outcomes and then play whatever it wants,鈥� he says. 鈥淲hat the parrot has presented is this beautiful, elegant solution for making vocal sounds.鈥� Human brains have similar vocal maps.

Long and Yang repeated their experiments on zebra finches (Taeniopygia guttata), which are not vocal mimics. 鈥淭hey have a single song that they learn,鈥� says Long. 鈥淚t鈥檚 about a second long, sometimes less.鈥� Perfecting it takes them months.

Unlike the budgerigars, the zebra finches showed no sign of a 鈥渕ap鈥� of vocal sounds in their brains. Instead, 鈥渁 zebra finch develops this really almost impenetrable code for its song鈥�, says Long. He says the budgie brains use a simple and intuitive system to produce their complex calls, while the zebra finch brains use a complex system to make something simple.

鈥淚t shows that the neural activity and associated vocal behaviour is closer between parrots and humans than parrots with songbirds,鈥� says at The Rockefeller University in New York.

鈥淎lmost everything we know about the detailed mechanistic basis of learned vocalisations comes from a few species of songbirds that sing relatively simple songs,鈥� says at Cornell University in New York. 鈥淧arrots therefore provide an incredible opportunity聽to study both the mechanisms and the evolution of complex vocal learning and production.鈥�

There are several possible reasons for budgies to have evolved mimicry, says at Cornell University. One is courtship. 鈥淭he females actually prefer males that have better imitation abilities,鈥� he says, and if the male loses the ability, 鈥渢he female becomes more likely to cheat on him鈥�. Budgies also have very dynamic social lives: 鈥淭hey form small groups for a few days.鈥� Once a group has become established, the members start making distinctive 鈥渃ontact calls鈥�. 鈥淧eople think it might be like a password for this group,鈥� says Zhao.

Other skilled mimics may have similar vocal maps in their brains, suggests Long: 鈥淢y very strong guess would be that other parrots have the same feature, but it simply hasn鈥檛 been explored.鈥� He also suspects lyrebirds, which are phenomenal mimics that can even imitate artificial sounds like camera shutters, have something similar.

In the long run, Long hopes that studying how budgies generate their sounds will help us understand speech disorders in people. Those who have had strokes often experience aphasia: the inability to call the right word to mind. 鈥淵ou鈥檒l reach for that word and it鈥檚 not there,鈥� says Long. 鈥淣ow we have a fighting chance of understanding what I think is at the root of many communication disorders that affect people in devastating ways.鈥�

Bonobos are quick to help a person who doesn鈥檛 know what they know, a sign that they can deduce the mental states of others.

The capacity to think about what others are thinking, known as , is an essential skill that allows humans to navigate their social worlds. It enables us to recognise that someone may hold different beliefs or perspectives to our own, underpinning our ability to understand and help others appropriately.

The question of whether our closest living relatives also have theory of mind has been hotly debated for decades. Despite some , non-human great apes seem to have some aspects of this capacity, suggesting it is more evolutionarily ancient than once thought. For example, wild chimpanzees that see a nearby snake, albeit a fake one, seem to call out to they know haven鈥檛 already seen it.

But we have been missing clear evidence from controlled settings that primates can track a perspective that differs from their own and then act upon it, says at Johns Hopkins University in Maryland.

To investigate this, Townrow and , also at Johns Hopkins University, tested if three male bonobos at the Ape Initiative research centre in Iowa could identify ignorance in someone they were trying to cooperate with, and then gesture to them to help solve the task.

On a table between the bonobo and an experimenter were three upturned plastic cups. A second researcher placed a barrier between the experimenter and the cups, then hid a treat, like a juicy grape, under one of them.

Free newsletter

Sign up to The Earth Edition

Sign up to newsletter

In one version of the experiment, the 鈥渒nowledge condition鈥�, a window in the barrier allowed the experimenter to watch where the treat was placed. In the 鈥渋gnorance condition鈥�, their view was completely blocked. If the experimenter found the food, they would give it to the bonobo, providing a motivation for the apes to share what they knew.

Townrow and Krupenye looked at whether the ape pointed at the cup, and how quickly they pointed, after the barrier had been removed over 24 trials for each condition.

They found that, on average, the bonobos took 1.5 seconds less time to point and pointed in approximately 20 per cent more trials in the ignorance condition. 鈥淭his shows that they can actually take action when they realise that somebody has a different perspective from their own,鈥� says Krupenye. It appears that bonobos understand features of what others are thinking that researchers have historically assumed they didn鈥檛 comprehend, he adds.

This simple yet powerful research gives experimental support to existing findings from wild apes, says at Durham University, UK. However, she warns that the findings may not apply to all bonobos because the study animals were raised in human-oriented environments. But that doesn鈥檛 detract from the results demonstrating that a capacity is there, she adds.

Indeed, finding this capacity in these three bonobos indicates that the potential exists within their biology and, very likely then, the biology of our common ancestor as well, says Krupenye.

鈥淚t suggests that our ancient human relatives likely also had these abilities and could use them to bolster cooperation and coordination with one another,鈥� says at the University of California, Berkeley. 鈥淏y understanding when someone may be ignorant, especially about evolutionarily critical information like the location of food, our ancestors could have used these capacities to communicate and coordinate more effectively with their social partners.鈥�

Some snakes seem to respond differently to their own scent when it has been altered, which hints that they have some form of self-recognition.

A handful of animals, including roosters, horses and cleaner fish, have shown signs of self-awareness in what is known as the mirror test. This involves putting paint on an area of their body that they can鈥檛 see without a mirror, such as their forehead. If the animal touches the mark when looking in the mirror, it suggests that they are aware that the reflection is of themselves, and not an image of another individual.

鈥淏ut snakes and most reptiles primarily interact with their world through scent,鈥� says at Wilfrid Laurier University in Canada. So he and his colleagues challenged them to an alternative, smell-based version of the mirror test.

The team members collected the scents of 36 eastern garter snakes (Thamnophis sirtalis sirtalis) and 18 ball pythons (Python regius) by wiping cotton pads along their skin.

They then presented each snake with five scents: their own, their own with a bit of olive oil added, just olive oil, one of another snake of the same species and one of another snake with a bit of olive oil.

The garter snakes performed more long tongue flicks in response to their own modified scent compared with the rest of the scents.

鈥淭hey only do long tongue flicks when they鈥檙e interested in or investigating something,鈥� says Miller, which suggests that the garter snakes can recognise when something about themselves doesn鈥檛 smell quite right. 鈥淭hey may be thinking: 鈥極h, this is weird, I shouldn鈥檛 smell like this.鈥欌��

Ball pythons, on the other hand, responded in the same way to all the scents. Garter snakes are much more social than ball pythons, says Miller, so it may be that social species are more likely to have self-recognition.

The findings are the first evidence of potential self-recognition in snakes, says Miller. 鈥淭here鈥檚 this assumption that snakes, and nearly all reptiles, are these sluggish, instinctive, non-cognitive animals, and that is definitely not true.鈥�

However, at the University of Salzburg in Austria questions whether this should be interpreted as self-recognition. 鈥淭his interpretation only becomes plausible if a correlation with social behaviour can be established,鈥� he says. Otherwise, it could be argued that some snake species are simply more inclined to interact with the experiment.

Dogs seem to understand that words represent specific objects, recordings of their brain activity suggest.

Although some dogs can fetch a wide range of different objects on command, few do well on such tests in the lab. In addition, it is unclear if dogs understand words as object names, rather than instructions.

To explore this question, at E枚tv枚s Lor谩nd University in Budapest, Hungary, and her colleagues tested 18 dogs from a wide range of breeds, including Border collies, toy poodles and Labrador retrievers.

Their owners chose five objects familiar to each dog. In the test, they said the name of an object and then showed the dog either the named object or a different object.

Each dog鈥檚 brainwaves were monitored via electroencephalography (EEG) to see whether there was a difference in activity when the dog鈥檚 owner said 鈥渂all鈥�, but showed a stick, for example, compared with when the word and object were the same.

鈥淭he idea was that if dogs understand the meaning of the words, their brain responses will differ between the presentation of matching and mismatching objects,鈥� says Boros.

The researchers found that the EEG signals were different when the objects didn鈥檛 match and the effect was stronger for words that individual dogs knew well. This is similar to results seen in humans and suggests that dogs know that certain words represent certain objects.

鈥淭he most important realisation of this study is not only that non-humans are capable of understanding words referentially, but this capacity seems to be generally present in dogs as well,鈥� says Boros. 鈥淭his study demonstrates that dogs may understand more than they show.鈥�

No breed appeared to show a greater language ability than any other, says Boros.

at the University of Adelaide, Australia, says the study adds to the knowledge of dog cognition.

鈥淚 think dogs both understand more and less than what we realise,鈥� says Hazel. 鈥淭his research shows dogs appear to make a mental representation of a word they know 鈥� for example a ball 鈥� which is not at all surprising to most dog owners who know how their dogs understand some words.鈥�

On the other hand, she says, many dog owners anthropomorphise their pets and attribute emotions and comprehension abilities to them that don鈥檛 exist.

鈥淒og cognition is now one of the most studied areas around the world,鈥� says Hazel. 鈥淚 love all the research on dogs, but would love to see more on other animals we live closely with 鈥� cats, rabbits, horses.鈥�